Modern Middle Manager

Change Your Mind?

2007-01-03T10:31:00.000-08:00

One of my favorite stories goes like this: after being given a very disturbing prognosis delivered in an offhand manner by a medical resident, a real doctor came in to examine my condition. He was far less flippant and weighed the risk of the resident's prognosis as highly unlikely. The resident, still eager to show off how intelligent he was, declared my condition to be the result of damage to a particular nerve. The response from the doctor was, "Are you sure?" The resident, slightly less confident, repeated his conjecture. Once again, the doctor stated, "Are you sure?" Even after years of management, I know when you're being given a second or third chance to get the right answer, so I helped the resident by saying, "Sounds like you need to change your answer." I rather enjoyed getting in that barb.

On Friday the 29th, I had a conversation with the senior manager of our biggest business unit (we'll call him BK for Big Kahuna) about how to approach his extremely important make-or-break the company and his career project for 2007. I gave him my perspective, which he initially found appealing. The following Tuesday I called him to find out if he had accepted my solution. The conversation went something like this:

Me: I was wondering if you made a decision on our approach to the Really Big Initiative.

BK: Yes, you were going to send me a recommendation.

Me: I was? {Note: This part should not have been voiced. Duh.}

BK: Yes.

Me: I thought I gave you my suggestion on Friday. {Note: See how I grow denser.}

BK: You need to factor in [thus and so].

Me: [Finally catching on.] I will have my analysis and recommendation to you by Friday.

To recap: I made an practical, appealing suggestion on Friday that I thought was all but a done deal. On Tuesday, it was turned into "you are going to give me a recommendation." But I already had. That was my hint. And so I wrote the recommendation up, giving weight to the "thus and so." And he had the answer that affirmed the approach he already wanted me to take.

That is how office politics is done.

Projects

2006-12-31T06:57:00.000-08:00

Sunday morning before a flight to Paso Robles and my mind turns to...projects. Specifically, keeping the project stakeholders coordinated the business unit manager doesn't think anything's moving forward. Out of a portfolio of 12 projects, we have one project that suffers from substandard programming and will be moved to our offshore team. The remaining projects are either 1) in development or 2) awaiting the requirements phase, meaning that the project owners aren't writing a specification, they want the imaginary BSA that the IT department doesn't have available to write the spec. That's a shame and it holds up our projects. However, it also means we'll keep Orange County's job rate low. So we're procuring a body.

In the meantime, the business unit manager (and also my boss) thinks that someone needs to be in a position to coerce both the projects group (not a PMO, mind you...more an application development group) and the business unit product manager to "move things forward." So I may see the development of a project management office (PMO) that reports to me, or him, or someone else and partially runs my IT projects group. Or the IT project group may report to the product manager. Or to my boss, who would prefer to be more hands-on sometimes and less hands-on at other times. I smell a reorg coming (again). My empire has both grown and been sacked by the barbarians before, so I roll with it. In the long term, the Visigoths find that managing really isn't to their taste and doesn't accomplish what they desire: agility and speed to market without planning. I do not think that is a worthy target.

Here's a Challenge for You...

2006-12-28T20:19:00.000-08:00

The COO of the company says that he has a major initiative that will involve a great deal of software development. Without actually starting the design and specification process, he wants to hire more programmer (specifically, software developers located at a subsidiary in India). Do you:

1. Calmly explain that it is unlikely that we can hire the right number of people before we even know what is involved in the creation of this application / suite?

2. Say, "Yes, sir," and try hiring a half-dozen developers who very well may sit on their ass for weeks and leave out of sheer boredom before development begins?

3. Try to find another route, such as hiring a well-known offshore development firm to start the project, and then scale up the Indian subsidiary as phases are completed?

You decide!

What Happened to 2006?

2006-12-27T07:28:00.000-08:00

This has been a tumultuous year for us. We've tackled multifactor authentication, established an offshore development group, reorganized the department, gained people, lost people and accomplished a few business initiatives along the way. I've made some people happy, frustrated others and finally proven once and for all that third party examiners are a cross between Homer Simpson and Satan. Homer Satan. That sounds about right.

On a personal note, I got married in April, became a private pilot in November and gained about 20 pounds during the course of this year. I look forward to making that disappear in 2007.

Driving Desktop Operational Costs Down

2006-02-19T15:13:00.000-08:00

What a droll title for a post. Next week my group is working on two intertwined projects in an effort to drive down our ongoing operational costs. I've mentioned one guiding principal in doing that -- server-based computing. We use Citrix's MetaFrame XP product at this time, presenting the entire desktop on a Citrix server. Our proof of concept over the first part next week will encompass the following changes:

1. Moving to the enterprise version of Presentation Server 4.0.
2. Incorporating the Citrix Access Gateway to provide an SSL VPN and customize the kind of access that can be had to internal network resources.

The goals will be to gain the advantages of PS 4.0 (better printing, CPU usage control, server load balancing at the software level) with the addition of the CAG's security options. I forsee existing costs going down because we eliminate the #1 and #2 problems our end-users have with the existing platforms (i.e., printing issues and the follow-me desktop). In addition, we will be able to provide a secure method to publish certain applications for clients that require a highly secure connection, slashing the necessity to maintain private networks to those client sites.

The second proof of concept going on next week is a test of Softricity's SoftGrid. We will be determining if software virtualization is all it's cracked up to be. If so, imagine no need to image desktops anymore. Take them out of the box from your favorite provider, patch, install your favorite client antivirus/firewall software, join the domain. That's it. This dovetails the server-based computing initiative by virtualizing the software running under Metaframe. Set up the basic Citrix server, join the farm & domain, publish the virtual apps for whomever logs in. Done. No more software conflicts. Disaster recovery procedures practically write themselves, similar to the ease of recovery we've experienced from virtualizing our servers. I'm sure that the devil is in the details (i.e., actually "sequencing" the application to virtualize it). However, if it is a a one-time major pain followed by the occasional patch, I can't see it will be worse than the current software deployment and patch management nightmare.

Goals & Rants

2006-01-06T16:19:00.000-08:00

On Thursday I laid out the 30 and 90 day objectives for my staff. The good news is that none of them looked like deer caught in the headlights. The bad news is that I didn't get a lot of questions. There was a lot of nodding, which means I'll have a private follow up for the people who didn't want to talk in public. Management also means reading between the lines.

Ninety day objectives remind me of Optimize, a magazine that I have a love-hate relationship with. There's usually one article that I find very useful and the rest are just a waste of time. Maybe with all of this deli.cio.us tagging and such, a group can smart tag the good articles so I don't have to waste my time looking for them. I want Cliff's Notes for every industry magazine with some kind of Usefulness rating from my peer group. That would be helpful.

Microsoft IT Organization

2006-01-02T19:34:00.000-08:00

Looking for benchmarks I came across this page on Microsoft's IT organization. What caught my eye was two things:

1. Microsoft's organization is broken into two pretty standard groups, infrastructure and application development (internal IT services is basically infrastructure as far as I can tell). This is similar to the ideal division drawn up by Baschab & Piot in The Executive's Guide to Information Technology.

2. They rely heavily on contractors. That point doesn't sound as stupidly obvious as you might think. Yes, I am well aware of their use (and abuse) of contractors. However, I think I've missed something important over the years.

Once upon a time, I thought I'd never use contractors because I believed that you should hire employees, mold them the way you wanted to fit the corporate culture, invest time and energy into them and nuture them into great corporate citizens. That sounds great (OK, it sound like naive horse manure) but it doesn't necessarily work out. What happens when project demand is not smooth? What about keeping up the employee's technology skills while still getting the work done?

The answer may lie in exploring contract labor. For the first eight years with my current company, we institutionally feared outsourcing and contracting. Two years ago we broke the outsourcing fear and have outsourced/hosted several important support applications. Last year we used offshore development for the first time, implementing a software project that paid for itself before it was implemented in its entirety. This year will be the year we break the contracting taboo. I have a small cadre of very bright people working for me who are just not able to absorb any more training without significantly impacting their ability to execute. I do not believe that you use smart people by working them to death -- the more people work 50-70 hours per week, the stupider they get. So it's time to supplement our labor pool.

As I see it, contracted labor can be used in two ways -- to augment existing staff and on a per-project basis. I have need for both. The downside of contract labor is that their accumulated knowledge walks off the job probably every 6-24 months as a new person rotates in. The upside is that a contractor can be dismissed for any reason without turning the department inside-out if he/she turns into a problem. Bad employees are time sinks and it takes a long time to get marginal employees out.

Spending Metrics

2006-01-02T18:28:00.000-08:00

I started in on a book called "The Professional Services Bible" by John Baschab and Jon Piot. Messrs. Baschab and Piot have written yet another tome weighing in at 500+ pages. The virtue of their books is that they don't waste time with appetizers before serving the meaty topics. Within the first three chapters they have given the outline for the book and delved into professional services organization, management and benchmarking. Much of the book looks like it can be applied to any business organization.

Benchmarking is one such chapter. Some benchmarks may not be indicative of company success and should be viewed with a dim eye according to Christopher Koch of CIO Magazine. On the other hand, where do you start?

Our company took 3 metrics from the META Group (now a part of Gartner). Those metrics are:
1. IT spending as a percentage of revenue (Koch's least favorite).
2. Ratio of IT support personnel to total employees.
3. IT spending per employee.

We have managed our company to within 10% of META group's banking industry average for #1 and #2, including operational costs (something Koch contends is not included in the revenue metric). #3 confounds me, because I'm consistently 20-30% over the benchmark. I assume it's because we're a small company leveraging a lot of technology. I believe it's because we leverage technology in a big way and senior management agrees. Part of my argument is showing the return for specific revenue-enhancing projects and how they've paid off in a short amount of time.

Looking around on Google, I'm hard pressed to find more (free) information on metrics. One article I came across from 2002 is from Baseline and includes several more metrics.

Koch is correct in that picking benchmarks for the right reason is important and that trust should not be placed in a number for its own sake. I remember hearing Zig Ziglar say once that you will manage your organization to whatever you're measuring. Ultimately I (and the organization) need to ask whether the IT department is worth the 1/13th of every dollar of revenue we get. If we don't contribute more to the bottom line by helping to bring in more money, peel away expenses or mitigate appropriate business risk then we're a sinkhole.

Change Management, Still Ugly

2005-12-20T22:49:00.000-08:00

After two weeks of our initial CM process, I have proclaimed the weekly change advisory board (CAB) meetings a failure, taking 90 minutes to review our changes. That's 90 minutes for the three people involved in the CAB.

I'm a practical person. I don't care what the damned examiners want, if it's become a ridiculous drag on our business then we need to do something about it. So I've revamped our process as well. Changes involve relying more heavily on the help desk system, tagging the tickets with labels so that running reports on different stages in the process isn't as time-consuming. Because of the limitations with our help desk system, we are essentially running four reports. We're working on simplifying that but won't have that ready for another couple of weeks at the earliest.

The toughest part? I understand that some of our other departments have had problems using digital signing to show that reviews had been done. I'm hoping I can avoid that problem, but won't really know until our next exam. Wishing myself good luck...

Change Management, Two Weeks Later

2005-12-05T23:06:00.000-08:00

It didn't take long for the recent round of auditors to tell us that the change management process I started envisioning two weeks ago needed to be implemented NOW. Surprise! I held an hour-long staff meeting explaining the process, answered some questions, let everyone know that this would be a work-in-progress and then started on the hard work: putting together the procedures and memo for the rest of the company.

The most important thing for the rest of the company to know, if not understand, how this new process will slow things down. They say change management is like brakes on the car, it helps you go faster. Bah. For a small department, it means more paperwork and more bureacracy.

There is a game being played by this whole regimen of examinations and regulations. The game is this: if you are a company that hosts systems and/or applications, at a certain scale you can devote people to creating and maintaining compliance. Small companies can't and will be forced to outsource to the big guys who can, if they can stay in the game at all. So our leading lights of legislative prowess create barriers to entry in the public financial service arena. Thank you, Congress, for screwing the small guy yet again.

Offshoring Investment Research

2005-12-03T01:15:00.000-08:00

Sometimes that Barron's subscription actually yields a good investment idea. Not because it helps me pick better stocks. Oh, goodness no. I'm a terrible stock picker, worse when I actually listen to someone else. No, in this case they had a brief sidebar article on a new trend in outsourcing -- sending investment research to India.

Our wealth management line of business has a razor-thin margin. To be honest, it's not very profitable and we don't have the kind of scale or retail distribution that makes it incredibly successful like a Northern Trust or Merrill Lynch. So what does that leave us? Making what we have more profitable. If we could save 20%-30% a year in investment research costs by outsourcing that internal function, we would improve our margins for that line of business by about 10 basis points. Does that make it worthwhile? If it turns a not-so-profitable product into a profitable one, yes. If it outsources a few IR jobs to keep several other jobs, yes. At the end of the day, competitive advantage and shareholder value win.

I've read pundits who predit that the CIO/CTO function will eventually wither away, either outsourced or subsumed into some other business function(s). That day will be loaded with irony for me. Hope I like it.

Mobile Data Security

2005-12-03T00:57:00.000-08:00

If you do any business in California, surely you have heard of SB 1386. While I appreciate the warning provided to me as a consumer in this age of identity theft, I have to admit that it really throws one more damnable monkey wrench into doing business as a small company (or in our case, small division). For years IT has been assailed for not adding to the bottom line of the business. With the onslaught of more laws, whether SOX or GLBA or SB 1386 or whatever well-intentioned piece of legislation gets through the state or federal legislature (or hell, even created by the courts), a relatively small IT department like ours gets the Hobson's Choice of either derailing existing initiatives to comply with the law or covering its behind by letting management classify the regulation as a small business risk. In this case, we're choosing option #1 because we know how careful our salespeople and front-office personnel are with laptops and handhelds. That thought keeps me up at night. Check out the timestamp if you think I'm lying!

There are some vendors I've found in this space: Control Break's Safeboot, PointSec, PGP Corporation, Utimaco's Safeguard and TrustDigital. I'm sure there are others, but at least I've seen some reviews on these folks.

What are the points of pain? Let's count the easy ways to for employees and vendors to lose data:

1. Laptops.
2. Handhelds.
3. Removable media (e.g., CDROM and USB keys).
4. Tape backups (really a special case of #3).
5. Failed hard drives.

This list doesn't even touch on third-party network access, although we restrict that pretty tightly. Failed hard drives we physically destroy before tossing. The rest are out of our control. Password protection is insufficient as to keep from reporting under SB 1386; only encryption provides safe harbor.

I sent the memo to senior management. I am awaiting their response.

Mo Politics

2005-11-24T12:53:00.000-08:00

One of the continually fascinating aspects of my position is the need to understand and master the company's politics. Like a twisted game of Survivor, it's about making alliances and passing tests (i.e., executing on projects). Unlike Survivor, alliances are multidimensional -- you make them with subordinates, peers and executives. Subordinates and peers pull for initiatives, executives push. That combination can be very effective. For example:

One nagging problem that's festered for about the last four years is who will take ownership of business processes related to the company's CRM platform. The CRM is mostly used by a single line of business from sales to support. However, no one is left to own, maintain or review those processes, many of which are admittedly out of date. In fact, one particularly enterprising individual is trying to get changes made on his terms by declaring everything he doesn't like as a "bug" in the existing applications. How do I neutralize this turkey while getting some responsible, broad-based support for owning the processes and responsibly changing them?

A little digging and I find that there's a task force consisting of members of this LOB from the support side. Good start. I have an unrelated meeting with a senior member of this task force and starting ranting about obsolete business processes and who will take ownership, help streamline them and then work with my group to implement them. Naturally, she offers the task force as a great place to do this work. Heh.

Not too long after that, I start a rant in an unrelated meeting with my boss about the turkey who is trying to classify everything as a bug (my boss isn't too fond of this individual, either). I note that it's time the LOB took ownership of the business processes and their end results and I suggest the task force. He says it's time he told the head of the other LOB to show some interest in their CRM and figure out what they want with it instead of just complaining. Good, good. I'm pushing the right buttons.

The story is still being written and it may not work out quite right. However, knowing your audience and trying to get everyone on the same page without force of authority or bribes is called politics, and its something all managers who want to move an organization forward should possess.

Change Management

2005-11-24T12:44:00.000-08:00

We are currently undergoing a couple of routine audits (SAS 70 II and SOX), and one of the holes being pointed out to us is our change management process. It is insufficiently auditable and controlled as far as the examiners are concerned. I believe we're pretty decent for a small division, but SOX 404 does not care. So how do you invent a bureaucracy?

First you start by Googling everything in sight and finding out that there isn't much out there that helps. Most of the papers I looked at were more theoretical than helpful. Software vendors want to sell a solution that's customized just for me. Hey, that's great. What do I really want? Back to square one. Of course, the companies auditing us would love to design a "best practices" change management solution that would cost us an arm and a leg and probably require a full-time person just to manage it. No way.

What am I left with? I started by stealing the change management flowchart out of Microsoft's Operations Framework. The next part is to revise it so it's a bit more streamlined for our 2.5 FTE IT operations group. Finally, after kicking the ideas around with the two senior guys on my staff (and the ones who will have to live with this the most), we've decided to model the standard change process and walk through it on paper. Like paper trading, so we can empirically discover the flaws before I declare it to be The Process by fiat. I prefer to learn by doing, and most of my staff does as well.

It's Hard to Post...

2005-09-11T20:01:00.000-07:00

...with a broken arm and heel bone. I've been out since 7/30 after a head-on collision totaled my car and put me temporarily in a wheelchair. If nothing else, my staff hasn't had me breathing down their necks for over a month and they seem to be doing pretty well keeping things running.

Wheeling and Dealing

2005-07-05T20:43:00.000-07:00

I am not always the best negotiator. Case in point: I've tried to pry away a contract employee from his employer the proper way, coming to mutually agreeable terms that nullify their contract's existing do-not-solicit clause. After four weeks of negotiating, the way I found to break this impasse was to finally offer a ridiculously low sum of money and bring in our COO to play the Last Word.

First Round (3 weeks ago)
Us: We want the contractor and are willing to do a deal. What do you have for us?
Them: Contract-to-hire lasting 6 months and a 15% buyout at the end. I have to maintain my top & bottom line, you know.
Us: Ouch. We'll think about it.

Second Round (2 weeks ago)
Us: We still want the contractor. Can we link other business we drum up to replace the loss of our business?
Them: No.
Us: OK, what about 3 months at rate X with a 10% buyout.
Them: Too low. I have to maintain my 20% margins, you see.
Us: I'll see what I can do.

Third Round (1 week ago)
Us (via e-mail): I've given a lot of thought to this. We really aren't willing to pay that kind of premium. Final offer: we will give you rate Y for 100 hours of the contractor's time and rate Z for 100 hours of some programmer time. Then you lift the non-solicit. That's it.

Fourth Round (today)
Them: You insult us by offering a ridiculously low amount. I have margins. What happened to 3 to 6 months?
Me: We can't afford anything more. With the amount you're asking, we can get an outside project manager.
COO: Look, we want to be reasonable. What will it take?
Them: 15% buyout.
COO: That's steep. What we offered (the 200 hours at rates Y and Z) was a premium of A. You want a premium of B. Can't we meet in the middle?
Them: We only offer our best clients a buyout rate of 10%.
COO: Then we'll pay for 200 hours of support at rates Y' and Z' to make it 10%.
Them: Deal.

Yes, we probably didn't get the best deal. On the other hand, at least my initial attempts weren't so far from where we settled. And the ham-handed negotiating at the end seemed to get them to move.

The Royal Scam

2005-07-03T23:08:00.000-07:00

On Saturday I was posing as your everyday yuppie, mingling with people at a blind wine tasting event (brief plug: we were using the Bagged Wine product to perform the blind testing and it's up to the job). One of the guys I talked to runs his own consulting business installing systems management software. No biggie. However, he told me about one of his buddies who consults in the server virtualization and consolidation space (you know, VMWare ESX server and all that). Apparently this guy charges $1000 per server for virtualization and consolidation projects. I realize the savings are immense, but if anyone is talking to consultants to perform this project please realize this:

YOU ARE BEING TAKEN FOR A RIDE.

Good grief, go buy PlateSpin's PowerP2V product and save yourself some money. Better yet, pay me $950 per server. I'd rather make that kind of filthy lucre.

IT Cost Allocation

2005-07-03T22:17:00.000-07:00

I don't believe in fate or higher powers or the will of the universe or any of that mumbo-jumbo. Sometimes it seems like Someone or Something is mocking me by trying to prove me wrong.

In my last post I mentioned that my company has two factions, led by the head of banking and wealth management, respectively. Both of these guys want to be the CEO of the company, and I believe I can describe their strategy thusly: grow their own line of business (LOB) and cash cow the other one. I currently report to the head of Banking who is also the Chief Operating Officer. I have had hints dropped my way that my department will wind up reporting to both LOB's and become a football between them in regards to IT expense allocation. The issue is that the Wealth Management (WM) LOB consists of about 75% of the company's personnel but really doesn't need the high-tech core infrastructure that we've created for Banking. Naturally, Banking does not want to pay 100% of those costs, especially if it is shared in any way with WM. Hence my dilemma -- how to charge back "fairly." I really don't have time for this crap. However, like I've said before, politics is ever with us and to believe that you can avoid it at the middle or executive management layer is folly.

So what to do? I have pondered that for the last three months as the writing started to appear on the wall. Within the last month I ran across a couple of resources that got me thinking. The first one is the Professional Services Firm Bible (PSFB), written by John Baschab and Jon Piot (disclaimer: I'm not done reading it). This book appears similar to their Executive's Guide to Information Technology, which was a comprehensive reference guide to IT organization, operations and strategy. The second resource is an article in CIO Magazine on how an electric utility structured their IT chargebacks.

What does this mean for us? I believe we can create a similar model as Southern Company while providing the level of service expected by a professional services firm. Banking's #1 concern is uptime. Wealth Management's #1 concern is lowering their cost structure. Therefore, my model for a shared-service chargeback would consist of:

1. A per-capita desktop charge. We would have a full desktop or thin client offering, depending on what the LOB wants to spend. This makes it less likely that I'll need to argue the value of thin clients over desktops yet again.

2. A per-incident help desk charge. Again, this could vary depending on the SLA required. Banking generally wants priority over WM. Great, as long as they pay for it.

3. A per-capita shared infrastructure charge. This charge would be for all of the shared services (network, server, software, Internet) used by the company. These charges should be high enough to pay for upgrades every 3-5 years.

4. Demand-based infrastructure charges. This charge would be for equipment and maintenance used specifically by an LOB. For example, firewalls used specifically for Banking would be identified and charged to them. Branch office expenses, such as firewalls, routers and data circuits opened by WM would be charged back to them. This would put choices such as T1's vs. DSL into their hands rather than have to fight with them to reduce the costs of data circuits for 4-person offices.

5. Demand-based project charges. This is my favorite and where the PSFB comes in. I want to provide a cost-plus structure in order to make our services very attractive to both groups, including selling them on what we can do. I have no problem being competitive with outside vendors and believe I can provide what the LOB's need at a much lower cost structure.

Items #1-4 are to fund the infrastructure and include maintenance agreements, upgrades, capacity expansion and the IT Operations personnel necessary to keep everything running. Item #5 is the one I'm really keen about. Not only would it give my department more visibility and accountability, it would give us an opportunity to run as a profit center. I know that is not an easy task, but nothing focuses a department as much as responsibility for its own bottom line.

A major benefit of running a department with profit and loss responsibility is the opportunity to reward the people within it based on their results. Compensation is a subject I'll review more in-depth, but on a simplistic level I believe the IT Projects personnel should be rewarded if they "bill" certain levels back to the departments and those departments are happy with the deliverables. The IT Operations personnel should be rewarded based on uptime and "customer service" factors.

In theory, this would provide the Wealth Management LOB IT overhead expenses of the fixed (infrastructure) and variable (help desk, desktop and specific infrastructure/project) variety. If they believe that their cost structure is unappealing, it will be within their power to lower it. For Banking, their cost structure is less important -- they are willing to invest in IT for a return which would be far more measurable than today. They also desire a higher priority and greater uptime than Wealth Management, both of which should carry a premium. However, they should not be penalized for the per-capita expenses.

Unless I've lost my mind, this looks like a win-win situation for us all.

Another Battle Begins

2005-07-03T22:08:00.000-07:00

From time to time I have documented the factions that struggle for supremacy in our company. Last week another potential contender for the CEO throne, the head of our investments department, was derailed from his ambitions. Instead of being in line for the kingship, he was "booted upstairs" and will now work directly for our parent company. In many ways this is a good thing, for his skills are in sales and marketing, not in management. I do not know for a fact whose hand was behind this, but I have a pretty good suspicion. Naturally, I support it. Why? Because our investments department's culture has not been assimilated into ours. They do not yet serve the role for which they were acquired, i.e., to service our clients. Instead, they want to be a boutique research firm. However, their cost structure is still a drain on our wealth management line of business. I expect to see further departures in the near future, most of which will probably be welcome.

The race is now down to two senior managers -- the head of the banking line of business (LOB) and the wealth management LOB. This sets the table for my next blog post, which will be about IT cost allocations and how I believe I can prevent my department from being a football between these two fiefdoms.

It's Tuesday Again???

2005-05-24T18:55:00.000-07:00

I'm completely baffled by the speed at which time passes. Especially as of late.

Well, today's technology will be Platespin's PowerP2V software application. This is a tool that is great if you want to virtualize your data center (P2V -> physical to virtual). Our data center is about 90% virtualized with VMWare's ESX server, so why would we need this tool? Because sometimes you need to grow your virtual servers, and this does a great job of resizing those virtual hard drives without screwing around with partitioning software in a VM. Was 10GB too small for your VM? Too bad unless you have a tool like this. It paid for itself in headache and hassle after about 8 resizings. It works great with Windows guest operating systems and supports Linux, but only one flavor -- Red Hat. Too bad for us because we're a Debian shop. That's about the only knock I have against the product. I requested that they pursue Debian support, although I suspect it didn't make it very high on their list of product requirements. Aside from that, it is a good, solid product. At least for Windows P2V conversions.

Technology Review Tuesdays

2005-05-17T18:41:00.000-07:00

I write this as I try to shake off a cold I picked up in Vegas. Blah.

I've never made a regular, recurring topic on this blog. Maybe having one will inspire me to write more. Today's technology is from SyMark, which makes an appliance for password escrow called PowerKeeper. This handy, dandy little device controls the generic admin accounts running throughout the network, including such platforms as Linux, Windows 2000/2003, MS SQL Server and Cisco equipment. Passwords are "checked out" from the device for a specified period of time and with or without dual controls on a per-account basis. As a financial institution tired of getting beat up about having generic admin accounts (something we're remediating this year), I like what I see.

Access is controlled by the interaction of systems, collections of systems, users and groups of users. One group can be assigned to request passwords for a collection and another group can be set to approve the requests. There are two "superuser" accounts on the box that override all other security, which may provider something of a security hole. It looks tedious to set up and the HTML interface is pretty simple. However, once set up it seems easy enough to maintain.

The box produces about a dozen reports showing information such as checkout activity, password aging (if you choose to let the device automatically change passwords according to customized rules), and about 10 other items I don't remember from the demo. Reports can be scheduled and e-mailed as HTML or CSV files.

I'm impressed enough to request an evaluation box. I understand that pricing is based on the number of systems that it manages, so hopefully they won't be too greedy.

Project Methodologies and Us

2005-05-16T20:54:00.001-07:00

We've been using Scrum for a few months now and it works great for software delivery. We've had mixed results with its use for systems projects. I'm not entirely sure if it's the methodology or the personnel, though I'm trying to get a handle on that. Is Scrum too vague for the people I have? Are the participants behind on their skillsets? Why is it that developers can take off and work on implementing requirements while the systems people struggle to find time to work on their projects? I'm not entirely certain. Hmm.

Nothing New about Projects

2005-05-16T20:54:00.000-07:00

Meeting deadlines is better than blowing them.

Personnel

2005-05-16T20:34:00.000-07:00

The toughest part of management is personnel. Everyone says they want to be a leader, blah blah blah, but leaders generally don't worry about the dotted "i"'s and crossed "t"'s. They are busy preaching change and strategy, leaving the broken pieces to be picked up and sorted by middle management.

Damn, that's some jaundiced stream of consciousness that has nothing to do with my point about personnel.

Three weeks ago I restructured our department and eliminated a position (aka "put someone on the street"). An uninviting task and the first person I've ever directly sent packing. I have another similar task for a probationary position who isn't working out. She'll be gone at the end of Monday. It seems like I'm on a losing streak with personnel lately. And every time I make that decision, someone's losing their livelihood, maybe their home, maybe their chance to make it big. It's an ugly decision made under cold fluorescent lights in a small office with stale air. If you stop "sharpening the saw" and keeping up your skills, whether technical or managerial, the axe may come for you. And yet it's a sound business decision, not just for the company, its shareholders, bondholders and customers -- it provides opportunity for someone else who is hungry and hardworking and fulfills the needs of the organization. That new person may grow and prosper or may get complacent and get left behind. I'm just here picking up the pieces. And sometimes breaking them.

On a somewhat more positive note, I did put a candidate through hell today in order to gauge whether he could become a systems engineer in our firm. I think he did well -- not only polished on a personal level, but able to produce positive results on the practical skills part of the interview. I didn't get any heebee jeebees (HR jargon) from his resume or the interview. I hope I've done my due diligence a little better than the last time.

An Article I Found...

2005-04-17T22:30:00.000-07:00

...coincides with my last post. In it, George Spafford discusses the need for solid processes so that the pace of change does not totally disrupt an IT organization. Bottlenecks must be identified and addressed. What are our bottlenecks? Hmm.

On Outsourcing

2005-04-17T21:51:00.000-07:00

There is a reason why attrition over time may be a good thing for a department. Though I might try to persuade, cajole, coach and even punish, at some point it just makes sense to change a department's culture by changing the people.

I never thought I would consider outsourcing certain skill sets in my department. A year ago I decided to outsource a development project because our existing developers were maxed out and we needed the project done. I'm concerned that my current, small staff can't keep up with the number of business initiatives that we have. Either my staff needs to continue to adapt and learn or risk being left behind. This is the same story major IT departments in America are learning -- there is too much to do, not enough people to do it and no way to hold back the tide of projects.

This leads to making tough decisions about what skill sets to keep and what to let go so that others can do it better. Everything is an evolution. For now, it seems best to keep as much of our systems administration and app dev in-house as possible, outsourcing things like monitoring, maintenance, some advanced network design and overflow tasks. Over time, is it possible that in our small division we will simply be project leaders, managing vendors and services shared with our parent company? I fear this to be the expedient outcome.

Recovery & Rollback & Such

2005-04-17T21:05:00.000-07:00

An article in Datamation points out the importance of recovery and rollback in change management. This is a topic near and dear to my heart. My department's story is one of growth over the last five years; growth in responsibilities and maturity. The maturity angle is still catching up. It can be difficult to do this with a group that wants to live in the cowboy days of yesteryear. The new hires get it, the "old guys" are still adapting. I try the "carrot & stick" approach with the occasional whiner holdout complaining about how this makes them more unproductive so they can't get as much work done. Wah.

Restructuring

2005-04-17T19:50:00.000-07:00

It's been a while since I last read The Executive's Guide to Information Technology, but apparently it has had a firm grip on my subconscious. I have recently submitted a restructuring plan to senior management to try to make my department not just more agile, but more careful as well. The results are to split it up into a formal projects group and operations group. The projects group will have dotted-line reporting to a project leader, which also creates a training position for IT management. The operations department will be concerned with uptime, infrastructure, daily tasks & break/fix.

Being a financial company, I (and our examiners) are concerned with change management controls. I am finally putting into place a test/development network that will belong to the projects group. The operations group will be a double-check on their quality before anything new is implemented in production. A high-level overview to be sure, but this should bring up our quality while giving me 1) a better successor plan and 2) change controls that are company-friendly and examiner-approved.

Test Network in a Box

2005-04-03T21:02:00.000-07:00

I think I finally found it. The answer to all of my test network dreams and examination nightmares -- the Test Network in a Box.

Let me explain: I'm usually getting heat for having a half-assed test network. I've not spent the time or the energy of my staff putting together a test environment that mimics all of the critical servers on our network. The half-assed version we've got now is cobbled together from obsolete equipment and totally separated from our production network, which is a blessing and a curse. Blessing because my staff can't accidentally jack up our production Active Directory domain. Curse because it takes forever to refresh the test network.

Along came Dell to make my life easier. They introduced the 6850, a 4U server with up to 32GB RAM, four Xeon processors and 5 300GB SCSI hard disks. I outfitted one to that specs in an order I just placed. Combined with an Intel quad-1000MT card and VMWare's ESX Server for 4 CPU's and our existing NetApp filer. Because it's a test network, we can probably cram 40 virtual servers on the box. Our data is mirrored to the test NetApp filer on a regular basis and the virtual servers will be in their own vlans. Voila!

Security in the Growing Wannabe Enterprise

2005-03-26T17:47:00.000-08:00

When most companies talk about enterprise computing, they mean Large Companies. Big Stuff. Fortune 500. We're a small company that's growing about 20-40% YoY. Still small, though. I don't think we've even hit Medium Sized yet. That's OK. What's not OK is that our desktop security strategy still looks a lot like what we had 8 years ago. That's bad. While our perimeter security has improved tremdously, we have a soft underbelly that needs to be protected. I've been looking at running a Scrum to deliver on what I want to see, with the following goals in mind:

1. Protect the desktops against spyware/adware/keyloggers/Trojan horses/etc.

2. Protect the network against unclean machines (and/or send them to a site to be "healed." Heh).

3. Get a single sign-on scheme implemented for internal & external customers. The proliferation of passwords is ridiculous. Two-factor authentication is necessary for internal users as well.

4. Depending on how #3 works, perhaps get some sort of escrow for generic administrator passwords so they can't be issued or used without dual controls.

#1 we're addressing with Cisco Security Agent. It seems to be pretty impressive. We currently have it running in "test mode" on 70% of our desktops and 100% of our Citrix servers. We're looking to it to stop malware, especially what Internet Explorer lets in the door.

I'm looking at using Cisco Clean Access to address #2, although I think the creative minds on my staff should take a look at other solutions out there as well. The same goes with #3. I've seen some interesting products from Passlogix that might work. On a more limited scale, SSO products for web-based apps like Netegrity Siteminder or other SAML-based solutions could be of interest. I'm not sure what we need here.

I came across an interesting piece of marketing collateral a few days ago. The company is Symark and they have a hardware appliance called PowerKeeper that's kind of interesting. It's an appliance that escrows generic admin account passwords for Unix and Windows.

I'd *reallky* love to see an identity management solution with role-based access control that costs less than an arm and leg, works on both Linux and Windows, and provides SSO with multiple back-end repositories, but I'm not expecting it anytime soon. What I've outlined above should be good enough for the next 3 years.

Monthly Update

2005-03-26T15:25:00.000-08:00

I started this blog as an attempt to do several things:

1. Vent.
2. Figure out what strategy and direction I needed to go in.
3. Think out loud.

My need to vent has lessened over the past couple of years. The senior management team factions have solidified and a truce exists over the use of my department as a football. I'm please about that. It is temporary, though, and I know that I am not committing enough time to meet with the leader of the "other side," i.e., our wealth management LOB. I'd better start soon because he appears to be the likely head honcho in 2 1/2 years. Time to put a reminder on the calendar.

Figuring out the department's strategy has been easier. Doing it within the constraints I have is the tougher part. Between security, compliance and operational issues, I find that we aren't dedicating enough time to those value-added projects that bring a return to the company. On the other hand, according to several articles (here's one) I've seen on TechWeb's News, most organizations spend 70-80% of their time and energy on daily operations. Although my budget may resemble that split, my staff's time does not. Given that added value is coming from the creativity of my staff (as well as from a pretty good helping of open source), it sounds like we're doing better than average.

The kind of well-meaning managers that have driven me bats**t over the years are those that can't leave well enough alone. Now I'm one of them, always looking to change something and make it better. I hope that the difference I bring at this time is that when I've made a change and created a strategy, I know that it will take time to achieve it. I have set our strategy for this year and, assuming we don't get knocked wildly off course, am done thinking out loud. For a couple of weeks, anyway.

A Good Week

2005-03-13T19:55:00.000-08:00

Last week was the kind I enjoy seeing: low key progress on a number of projects, some creative thinking and, best of all, nothing went wrong.

One of our major applications, a client reporting app, is about to undergo a radical overhaul (understatedly called v2.0). This application demonstrates several great features:
1. It was based on an existing open-source project running on Linux using a PostgreSQL backend.
2. It was created by a single person on my staff.
3. It makes the clients really, really happy.
4. We can deliver updates rapidly.

This is one of our big client hooks -- they love the functionality of this app, although it grew in a way that requires taking a look at rewriting it with new technology. There are a few new changes we're looking at. One is the doubling of the number of developers on the project, another is using the MySQL cluster instead of PostgreSQL. The new design is going to focus on high availabilty, a more rapid and flexible history search and a number of improvements to the UI requested by our client base.

The best we can do for our clients is offer them as many of the features as they request, as fast as we can, with the greatest reliability that we can. This project is our best example of that.

Back from Edmond, OK

2005-02-26T15:28:00.000-08:00

I was sent out on a near last-minute trip to Edmond, OK, to perform an IT examination of sorts on a possible merger target. I came back with a fever and body aches. The best part is that I had a chance to schmooze with the president of the parent corporation, impressing him with my hacking cough and draining sinuses. Good deal.

Must get back to bed now. Feel weak. Blah.

Duck and Weave

2005-02-01T16:43:00.000-08:00

Part of the art of the middle manager is to duck and weave around the political games played by senior managers. I am apparently a part of three such games running currently:

Game 1: Defining business processes for the investment division. What's the goal? Damned if I know. The investments guy who is supposed to work with us to define his processes doesn't want to step on any internal toes and knows that the senior manager who assigned him the task won't advocate any changes that might involve other departments. My boss has decided that if that department won't fix their processes, he'll do it for them. I have been personally recruited to help implement his vision.

Game 2: Breaking up the company. The goal here is to report the financials of the company's lines of business separately, so that my boss looks good and his rival in senior management looks bad. The bad news is that if the political control of our company fractures, my department (and thus, me) gets exposed as a football between the two rivals.

Game 3: The biggest game of them all, between some senior managers in this company and some in the corporate parent and affiliates. We want to process a majority of the sister division bank deposits and there is at least one corporate "made man" who does not want us to do that. It has devolved into rumors, lies and backstabbing already. Our biggest leverage is that we make the clients happier than their current banking relationships (which are generally determined by the "made man"). I'm part of the latest charge to make us more vital to the sister divisions through some application interfaces.

No manager can exist in a corporation without some political skills. After all, politics is just the art of getting what you want from other people. It involves making deposits in emotional bank accounts, doing favors, generating good will and then calling upon those connections to move people, departments and companies in the direction you want to go.

So what about the games above? Number one will probably leave me tainted with some bad will in the investments department. The senior manager of that department acts as if he is powerless and weak. The leverage in that department is in the form of its chief investment office and one particular analyst. Persuading them of the overall benefits they will have if somebody applies some discipline to their processes is going to be a tough sell.

Number two is going to require some dancing on my part. If I can get both rivals to agree on standard metrics for prioritizing projects and maybe some chargeback scheme, there is a good chance that maintaining the peace will be reasonably easy. Chances are, though, that the senior manager who is not getting attention now will want his with interest. I suspect that an IT Steering Committee will need to be created at that time. I really just can't wait to broker the competing interests of those two. Especially when I will continue to report to one of them. It is a 1 in 100,000 chance that my reporting would change to the CEO. And I'm not sure that would help.

Number three is where we'll earn our money this year. My part of this needs to be executed solidly. We are trying to take a banking application and tie it into another division's workflow application while maintaining control of both our source code and the disposition of the deposits.

That's the game plan. Any questions?

Allow Me to Vent

2005-01-25T22:53:00.000-08:00

I am not in a good mood. It gives me heartburn to know that, after I've been here for almost a decade, there are still people here who are more willing to give the benefit of the doubt to someone they've never met than the department that continuously delivers solutions to them.

We're currently working with the e-mail admin for a business partner who is having a problem understanding that we do not send and receive e-mail from the same gateway. Besides not grasping the concept, he's having trouble with bridging the communications gap with my staff. What bugs me the most, though, is that one of our VP's just refuses to believe that it isn't anything other than problem with *our* systems. I have explained the issues point by point, demonstrated the near-total inability of the e-mail admin to communicate using e-mail, and the story she gives the Executive Vice President is that it must be our fault. We don't know what we're doing. Yeah, I've got that in e-mail.

Normally I would write this off. But there is something disturbing about a person's unwillingness to understand the issue, whether the VP or the e-mail admin. My goal is plain -- to get them both to acknowledge reality. I will have my staff continue to discuss the situation with the e-mail admin again. I intend to make him understand our setup, acknowledge how it works via e-mail, affirm that he will not adapt his systems to meet the needs of a Fortune 500 company, forward that on to the EVP and stick it in the face of his VP. Not that I'm spiteful. But I do expect a little gratitude and benefit of the doubt.

Do not screw with your CTO and defame my staff when you don't know what you're talking about. You'll be sorry.

Let's See How Agile We Really Are

2005-01-25T22:40:00.000-08:00

I had the pleasure of introducing two new projects to the staff today. Two new projects will be defined tomorrow. In addition, we have three outstanding projects running for at least another month. This week lays out how agile our department is really going to be.

Of the seven projects that will be running, two of them add revenue. The other five are compliance- and security-related. Of those five, four are related to some Sarbanes-Oxley cleanup. So I will wind up running my seven staff into the ground primarily to satisfy regulatory requirements, which seem to get tighter every year. Yikes.

The good news is that, with our new project framework, I think we'll get more of these accomplished in a shorter amount of time than we used to.

The First Big Project of the Year

2005-01-23T20:45:00.000-08:00

We're going to automate the entire process our sister 1031 exchange division has for creating and liquidating accounts. That division already has the workflow software, our company does the accounting and produces bank statements and another division perfoms daily reconciliation. Starting this Wednesday, all of those needs get addressed by the three divisions. Within 6 months, we will integrate all of this software and reduce the man-hours it takes to perform 1031 transactions almost in half. Throughout the entire enterprise. Naturally, our division will take a little part of the deposits to pay for it.

Sometimes companies try to move so fast that they have to scale up in a ridiculous amount of time, costing great sums of money. Sometimes they move so slow that opportunities are missed and potential sales lost. We've hit the sweet spot where everyone's ready for this...and it's going to be a great winner all around.

This is the kind of stuff I love to see!

Knowledge Made Real

2005-01-23T20:23:00.000-08:00

My last post lied. We did have some additional ripples from Tuesday's problem that one of my developers and I worked on until this morning. Gah. What a week.

The crux of the problem is "knowledge made real," i.e., what happens when you think you know something but it is working through the actual problem that exposes you to the limitations of what you learned and remembered. We have a number of gaps in the knowledge transfer that was supposed to occur on the afflicted systems and I'm irritated. Nevertheless, it's important to note that the last iteration of knowledge transfer on these processes reduced catastrophic downtime from 4-8 hours down to 30 minutes. That's good. But not great.

The next iteration is to reduce the time it takes to fix inconsistencies that occur with the data. To do so, it is necessary to understand and document the process in a way that can be quickly grasped. The last developer hired into my department surprised me by being good at documenting processes graphically. This is going to be a good, er, better introduction to her of the systems she's going to be responsible for. Now if I could have gotten things done right the first time...

What A Day

2005-01-18T22:35:00.000-08:00

Today started out with a banking system bug that spit out the wrong numbers for some clients to reconcile with. From there I got called into a meeting with our examiners and gave them an earful. Their examiners on site were, in some cases, half-assed. Finally I wound up my day grabbing lunch and pulling my developers together to form a game plan to fix the bug discovered earlier in the day (yes, it took that long to research it). Except for an hour break to have a beer with an ex-employee who has moved on to bigger & better things, it's been bug, bug, bug all day. We finally corrected the numbers and produced the right tonight. It's been a long day.

Preach It, Brother

2005-01-12T13:13:00.000-08:00

Once in a while I run into a pretty good screed. This one from a psuedonymous Fortune 500 CIO is a beauty about the ambiguity of the statement, "IT must align itself better with the business." I agree on most counts that this commandment is a Red Queen statement, the words meaning what you want it to mean. What is the easiest thing for IT to do that shows value to the business? By putting results on the bottom line. What are three easy ways of doing that?

1. Cut costs.

2. Add revenue.

3. Retain profitable existing clients.

Cutting costs should be done intelligently, usually through automation. Sometimes cutting costs can be achieved through renegotiation with vendors or a revamp of the IT infrastructure. We've done both over the years.

There might be numerous ways to add revenue to a company. The trick is doing it with a return on investment. What good is adding $100,000 a year in banking fees if you spent $1,000,000 to do it with annual maintenance fees of $200,000?

Running a PR Machine

2005-01-07T14:52:00.000-08:00

Rayne makes reference to the perception that internal clients have towards the IT department and whether I'm measuring the metrics that they care about. The answer is, "Sort of." We're a small company, and in a small company the end-user experience can be found pretty quickly by listening to the grapevine. What's the scoop? We're good. People are generally happy with the desktop experience. They aren't losing their documents, their systems aren't crashing (except in the case of hardware failure) and even the recent CRM upgrade hasn't really sent anyone over the edge. We've got a really positive image so far.

What about the back-end, where it all happens? Mostly good. We had approximately 46 hours of unplanned downtime during the year relating to our banking LOB. Fortunately, much of that impacted a subset of clients rather than all clients. It still counts against us, but at least it wasn't 46 hours of 100% failure. The way we calculate things around here it puts us at about 3 9's of service. Not great, but not awful, either. And no outage lasted greater than 2 hours. The average was about 15 mins.

Generally speaking, our internal and external clients don't notice us as much anymore. And I think that's great. Moving forward in 2005, we are going to take on training. It will turn out to be more than that, but the best part of training is the touchy-feely aspect of it all -- I have someone on staff who will be acting as the IT department's ambassador and probably doing it well. He's a people-person type of guy. The perception in the field will be that we care. And considering my budget, the amount of money we'll be spending for this is minor compared to the additional goodwill I expect to see us garner. So 2005 should be a really, really good year on all fronts.

2004 In Review

2004-12-29T17:53:00.000-08:00

Now that I've got some energy & time, I'd like to commit my thoughts on the year in review. Yes, there are two days to go...who says I can't be early?

What did we do right? How did the efforts of seven people and the spending of $2 million on IT benefit our company? Did we waste it? Did we accomplish anything? Are we in a better position this year than last year? Have we aligned ourselves with the business? There are many ways to view our results. Here's a report card for our company from a spending and project management perspective:

Where the Money Went:

Personnel Costs (incl. contractors): 30%
Leased Equipment (3 yr. lease, 10%): 29%
Maintenance: 11%
Outsourced Imaging Systems: 10%
Voice & Data: 8%

What We (Mostly) Bought:

1. Upgraded network capacity & redundancy.
2. Additional & upgraded hot site equipment.
3. Upgraded desktop equipment for merged division.
4. Additional physical servers.
5. Microsoft licenses for new servers and merged division.

Why We Bought It:

1. Upgraded network to continue to provide uptime to banking clients. So far, so good -- we've been solid and banking has added numerous new clients. Revenue for that line of business is projected to be $17 million this year. Deposits are up to about $300 million.

2. Hot site required several upgrades to keep pace with growth of production site. I hate spending the money for idle equipment, even with most of our servers virtualized. However, better safe than sorry.

3. We acquired a sister division and found that their desktops were about 4-5 years old and unstable. We fixed that. We also virtualized their servers and brought them into the fold -- that required #4.

4. Additional physical servers were added partly because of the merger and partly because of additional projects, mostly banking related.

5. Oh, and we found out that the sister division could not account for most of their Microsoft licenses and got stuck holding the bag.

Supporting the Business:

1. Infrastructure is a big deal, not for a solid return I can point to when things go right as much as the damage I can point to when things go wrong, such as slow client access, lost connections and unavailable resources. Am I spending too much money on infrastructure? Not when measured against the likely potential business risk.

2. Much of what we accomplished this year to add revenue was done with open source software riding on near-commodity hardware. I expect to see us continue to build most of our internal applications with LAMP and J2EE using clustered middleware provided by JBoss. The return on those applications has been fantastic. Now if open-source grid computing would get here already...

3. The upgrade to our CRM platform will put us in a better position to integrate data from several applications and display it in the new, web-based Onyx. However, I suspect the first such requests will be frivolous monkeying with the interface rather than any sense of getting "one true view" of the client. We'll see if I'm right.

4. A hidden expense in the budget is my department's staff and how they are deployed. My goal is to continue effectively reducing operational costs (both man-hours and money) without sacrificing the smooth running of our infrastructure. Personnel is then freed up to pursue projects that add value to the organization and implement those projects in less time. I am not yet satisfied with the gains we've made so far, but we have definitely made progress over the beginning of 2004.

What Could Be Better:

1. The cost of our outsourced imaging system is looking excessive. The current percentage includes some post-conversion programming as well as operating costs, but I'm concerned about the amount of money spent. I'm not sure we're seeing the return we expected vis-a-vis our old platform. On the other hand, the quality of indexes and scanned images appears to have improved. This requires some wait-and-see over the first two quarters of 2005.

2. As I alluded to above, I think we should be completing more projects. This is partly because the operational duties of my staff have increased as we've increased the number of servers, networking equipment and other infrastructure components. I think it's also due to knowledge gaps on my staff that I'm attempting to backfill. There are some interpersonal dynamics within my staff that need addressing as well, although the dysfunction is not so great as to warrant radical changes. It's similar to how drag affects an aircraft -- it slows down flight but doesn't cause the plane to crash without other influences.

Report Card:

Overall Project Execution: B- (still have some lingering one-man projects)
Overall Project ROI: C+ (cost savings & added revenue were OK)
Total Spending: B (within 15% of industry standards)
Overall Grade: B-

Concluding Thoughts:

2004 was a good year in which we got a lot done and executed on some reasonably large team projects that cut costs and added to the bottom line of the organization. Unfortunately, some one-person projects still linger, leaving us a a little behind the curve. Costs are generally within industry standards with some categories looking a little worrisome.

For 2005, I expect to see larger projects realized, especially those that contribute to the bottom line of the company.

Dedication or Death March?

2004-12-19T23:12:00.000-08:00

They're sooo close to being done...

My staff is finishing up a three day upgrade to our CRM system. They have worked Friday night, all day Saturday and are trying to wrap it up tonight. They've already blown past their rubicon, the go/no go deadline I told them to set for themselves.

"If we could only get one more bug fixed, it would be done," they say.

And I'm letting them do it.

Management exists to further the aims of the organization. Sometimes that means taking a risk that your bright, well-trained and motivated team will totally blow it by obsessing with the problem, creating a tunnel vision that will swallow them up. Yet I do not want to pull the plug on their efforts, which have been substantial. So I am waiting for them, checking up on their final push and encouraging them, even while I listen to make sure that they aren't in a fog of weariness, making bad decisions. It's going to be a close call. The rollback plan will take about an hour to perform, so that's good news.

What will they take from this experience if they succeed? Hopefully confidence in their abilities, tempered with a tongue-lashing about trying a death-march with every project. It will be to their credit that they overcame a number of problems during the actual upgrade that they did not find in their trial runs. They performed pretty good triage on the system problems from what I can tell, although I need to find out more of the details.

What if they fail? It would be a heartbreak for the team, but there would be some good reasons for it:

1. The test network that the upgrades were done on was not 100% like the production network. It turned out that some Microsoft patches applied to the production network caused some early failures and set back the project by hours.

2. The migration of data didn't turn out as smoothly as it did on the test network. I don't have as much information on this as I'd like, but the audit of converted data could have been more complete.

3. By being overtired, they aren't going to be able to support the product very well tomorrow morning.

Scrum Starting to Show Results

2004-12-12T20:51:00.000-08:00

We currently have one project finished, two more starting and one almost wrapped up that have been more or less managed with Scrum. The first project was about 30 days long, the second will last 90 days and the other two are probably 30-60 days in length each. Here are some of the things we've found:

1. The group really works together well, although there have been some communications issues. As long as the same team works on more projects, those issues get resolved and their teamwork evolves. I expect bumps along the way, but it's good to see people work together.

2. The product requirements are getting defined pretty clearly but I'm still unhappy with how well the teams define the QA portion of the project. I think QA needs to be an explicit product requirement.

3. Nothing has gone terribly wrong to disrupt a project plan, which is the real test of a methodology in my opinion. We'll see what happens with the latest project -- its rollout is scheduled for next Monday, which will pretty much complete the Sprint.

4. Because of the size of my department, everyone is committed part time to one or more projects. This really changes the dynamic of their daily operational priorities vis-a-vis their project priorities. It's been an interesting management problem.

That's it for now. Also, I've introduced the COO to the idea of a project portfolio. These projects and their total times will have a total cost associated with them so I can track their total expense and ROI, if any. I foresee needing that towards the end of next year in case we compete for resources with the money-sucking line of business.

Soap Opera

2004-12-12T09:33:00.000-08:00

You know, it just occurred to me that one reason I haven't written as much lately is that the soap opera factor at work has been dialed down quite a bit. Not that there isn't tension, but senior management isn't changing their tune anytime soon so that stalemate remains the same. We have some tension between us and certain executives in our corporate parent, but we're also looking to maneuver around that in the next 6-12 months. Maneuver can be the best weapon wielded in a corporate fight. Once we've deployed and integrated the product that transforms that fight, I will talk about it here.

Charting the Right Course

2004-12-12T08:20:00.000-08:00

One day I'll leave my position at my current company. I've been here for almost eight years, so I expect I'll get an interview question along the lines of, "Why did you stay at Company X for so long?" I already have a stock answer. It's an answer that came to me when I watched the dot-com companies fail, combined with the experiences of a job-hopping hotelier I once met. It is:

"Because I had the opportunity to see the consequences of my decisions over the long term."

If your career is made up of 3 year snapshots at a variety of companies, I don't think you can get an appreciation for the long-term results of your actions. Over my tenure here we've gone from a 3-server computer closet to a 60+ server datacenter, we've client-facing web apps for our wealth management clients, we've started a commercial banking line of business practically de novo and the IT department has quadrupled in size with all of the usual attendant issues. The company has grown, shrunk and grown again. Senior management has turned over a few times. New government regulation has been written and applied to our business.

Each time a change occurs, decisions get made and their consequences ripple over time. To face your own legacy, overcome your own mistakes and adapt the department you grew shaped to new conditions strikes me as a challenge worth tackling.

Gobbledygook

2004-10-29T15:06:00.000-07:00

Gobbledygook is a word coined to describe meaningless jargon. I found a great blog today that I thought was an inspired example or a satire. You be the judge.

Compare that to the entries made in this management-related blog.

Our Forays into Project Management

2004-10-29T13:57:00.000-07:00

We have a number of projects in our current portfolio. Most of these are the "usual" methodology whereby people make up steps and milestones on a Gantt chart, stay the course regardless of the obstacles found along the way, occasionally claim success when none in fact occurred and otherwise blunder on towards uncertain completion months after the project end was supposed to have occurred.

As a reaction to the lack of results we've achieved, I've been looking at Agile-type project management. Specifically Scrum. We've modified Scrum a bit because our IT-related projects are typically 6 months or less in timeframe and the people involved are usually part-time on any single project. Of our first two projects done "Scrum-style", one is almost 100% complete as the end of the Sprint approaches and the other will have its first Sprint Review/Retrospective next week. It is likely a 3-4 month project. The good news is that everyone appears to be contributing positively and the work is getting done at a rapid pace. I am still concerned about the overall quality of the second, lengthier project. What I am not as concerned about is reaching the end of a Sprint and having 30 days wasted, because 1) I get weekly updates on the projects from the Project Leader/ScrumMaster/whatever, and 2) I meddle a little. I don't change the requirements but I make sure they're thinking about issues. I don't want a train wreck for the sake of a learning experience. So far it's coming together pretty well.

What about more process-oriented models, like the Project Management Maturity Model that Rayne commented about? I'm still figuring out how to integrate that into our project portfolio without introducing more bureaucracy than necessary.

Quiet Day Today

2004-10-29T13:24:00.000-07:00

It's the end of the month, which means two things usually:

1. Our wire transfer volume spikes.
2. It seems like half of our company took the day off. I can best tell this by the number of help desk calls I hear...almost 0 so far today.

So I turn my attention to other things, like reading papers on user provisioning or thinking about the Disney suit. Is CEO/executive pay completely out of whack with corporate performance? The answer seems to be "maybe."

According to Watson Wyatt, a management consulting organization, executives whose pay is linked to stock performance by stock ownership rather than options show a positive correlation between their pay and the company's performance. Well, duh. However, companies like Disney that assign golden parachutes and stock options of unbelievable proportions for half-assed work become the poster-child for bad governance. Paul Volcker, former Fed Chairman, is concerned with corporate governance in general. When salaries get out of control in sports the usual answer is a salary cap of some sort after prolonged paralysis. What is the answer for fixing executive pay?

User Provisioning

2004-10-29T13:02:00.000-07:00

You know, I really didn't have any idea there were tools to perform user provisioning. I should have guessed -- there are software tools for everything. I'm half surprised someone didn't come out with a user provisioning appliance for enterprises so that it would go faster with dedicated ASICs and everything!

User provisioning -- why would a small business like ours need something with a fancy title? Because we're audited and, in these days of Sarbanes-Oxley and Gramm-Leach-Bliley, we have to make sure the right people have the right access to the right data and that the wrong people don't. Then we have to prove it. Proving that we have controls in place takes longer than actually complying with controls.

Thus comes interest in tools that help us prove we actually do what our procedures say we do. Software that provisions user accounts according to roles is a nice-to-have feature that prevents the occasional brain-fart. Software that can generate reports showing that we actually do that is valuable. I'm researching some possible solutions from the vendors mentioned below:

abridean
COSuser by OSM
Courion
ID-Synch by M-Tech
Securic

What's in Your CIO?

2004-10-11T18:51:00.000-07:00

On cio.com is an article about the nature of the CIO role. Is it more of a strategic position working with the executive board? Or a low-cost support function reporting to the CFO?

Hmm. Quite the false dichotomy in our case. We have a third way -- both. We work hand in glove with one line of business, participating pretty fully in its initiatives. Our other line of business generally gets the low-cost support as we continue to drive down IT expenses by standardizing, consolidating and virtualizing. We provide them a stable computing environment and tactical project support but not much more. This situation, although dictated by politics, has shown some important fruits.

Presuming I had to choose investing the company's dollars between the two lines of business, which would I pick? The one we're already investing in. I can measure the return for most of these projects. The ones I can't project an ROI for are at least justified by mitigating risk. I can help the other LOB reduce costs and maybe improve productivity, but I can't give it the services necessary to compete in the marketplace. The competitive landscape moved a while ago for it and, for various reasons, it never caught up. Without a significant investment I don't think it ever will.

I have no vision, no grand plan for how to help our other LOB. Like a sick friend, I can at least make it comfortable.

It's Budget Season Again

2004-10-10T22:50:00.000-07:00

Another fourth quarter, another budget. So what are the initiatives for 2005? A short list of the definitive initiatives:

1. Move to a new building. We're packing up most of our divisions located in three buildings and consolidating to a new building and a new data center in the middle of next year. Fun!

2. Security. Lock it down and control it because compliance and governance are the watchwords. There is no perimeter anymore and make sure everybody's accountable for what they can access and how they do it.

3. High availability. Three nines not enough...go for more.

After that it gets a little hazy, especially in terms of corporate strategic planning. I have no idea what the lines of business are planning on, even though the head of banking is my boss. We're going to pick up business here and there with no overarching goal. The wealth management LOB is probably even less clear and, for the first year ever, I'm thinking of passing on meeting with them. No point really, not when it's impossible to win trying to service them strategically. The investments department appears to be its best to play by their own rules and will probably get flattened by someone.

Nevertheless, we sally forth...

Still More...

2004-10-10T22:37:00.000-07:00

...OK, not this time. I'm still absorbing the latest from Agile Project Development with Scrum. It appears that Scrum may fit the model of project development that we use here -- lots of unknowns and complexity crammed into short deadlines.

Rayne, I know you've mentioned using CMM or the like for strategic projects. Would you mind sharing an example in the comments? I'm curious about how to meld the two styles of project management.

More on Change Management

2004-09-19T18:28:00.000-07:00

Reader Rayne sent me an e-mail (thanks again!) about her experiences with project management. These thoughts led me to a larger set of questions.

There are evidently at least a couple of project management perspectives (I refuse to say, "paradigms") floating around: those that are very process-oriented such as the Capability Maturity Model and those that are focused on quicker results, such as Agile/XP/Scrum. The context of the examples are software development, but the ideas can be extrapolated to any kind of project. So which is best? And for what? And how do I take these processes and apply them to a department of 6 people?

I've got a lot of thinking to do.

Change Management

2004-09-11T11:21:00.000-07:00

One of the areas that I've been completely ignoring is the creation of formal change management processes to govern our systems & network. Although my existing method, commonly known as, "Don't blow anything up during production hours or I'll hand you your head," seems to be working OK, it's not scalable. Our SAS 70 audits have identified this as a potential weak spot vis-a-vis industry practices. So, maybe it's time to rethink this.

I have been toying around with the division of my department into two parts -- Operations and Special Projects. Changes, aka Project Requests, would first go through a Project Leader for prioritization and scoping. The project document would be reviewed by the Operations department to make certain it meets the established standards and practices for the company. Revisions get sent back. Once OK'd by Operations, the Project Team performs the implementation which is audited once again by Operations.

The Special Projects group becomes the rapid implementation team once they have been given permission. I see them as performing a number of their tasks outside of regular hours, so flexibility and rapid assimilation of information are key.

Besides being the watchdogs of the computing infrastructure, the Operations department will be responsible for defining standards, maintaining existing systems, automating regular processes and creating security policy. This is a detail-oriented group with a focus on uptime and stability.

I've been drafting some job descriptions to fit into this new scheme. More on that later.

Cisco Solves My Problem. Sort Of.

2004-09-06T19:33:00.000-07:00

In answer to my last post, Cisco is coming out with their Network Admission Control solution to control untrusted hosts accessing a trusted network -- think vendor or visitor dropping their laptop onto the network. If I'm lucky, I'll be going to their offices in NoCal to learn more at the end of September. Stay tuned.

What's Next?

2004-08-20T21:28:00.000-07:00

So what else is on tap for network security? I have some wish lists:

1. I want to stop those damnable keyloggers (if they do happen to get in) from sending out data through the firewall.

2. I want to keep unauthorized desktops & laptops from being plugged into the network -- no ID, no link. Period.

3. I want to put our servers behind a firewall and control traffic.

4. I want this to be easy.

Yeah, I know -- good luck, especially making #2 easy. I see too many vendors in too many locations plugging in their potentially infested machines onto our network and think to myself, "Which one is going to bite us?"

I'm the Most Hated Man at Work

2004-08-20T20:47:00.000-07:00

Starting on Thursday we implemented web content filtering. I really expected wailing and gnashing of teeth, but instead got mostly muted irritation. I made certain every request went through the Help Desk software we have, captured forevermore in a database. Requests for urbanlegends.com, among others, did not go without remark. Please cite for me the exact business reason you need access to mypostcards.com again? Ah, I see. Sorry, no.

It took a few people a little reminder of why we implemented this in the first place. I see that a Download.Ject variant has been created to take advantage of more IE vulnerabilities. Or, to put it more succinctly, our content filtering is not about shaping people's work habits -- it's about protecting our network.

On a more amusing note, I figured out how to make rules on the ProxySG 400 complicated enough that it would reboot itself every 10 minutes. The real trick is to have a rule that examines the URL and compares it to a whole bunch of regular expressions, AND statements and NOT AND statements using more regex and various SmartFilter categories. That caused it to chew on its brain.

The good news is that most people took it in stride. My initial memo that I sent out about a week ago cited several reasons for the security measures, referring back to the CERT vulnerability list for IE and several recent articles on the Trojan horses popping in.

As the CTO of a regulated entity, I take the security threat very seriously. I do not think we have the option of relying solely on our antivirus scanners and we cannot switch from IE due to the websites we need to visit for our daily operations. This is the best solution for us. For now.

Nothing New

2004-08-12T20:35:00.000-07:00

It's been almost three weeks since my last post and I have to admit that nothing new has occurred. So where are we with our projects? A quick summary:

Revenue-Generating:

1. Creating web services for one of our clients to open & close bank accounts automatically. This is our capstone project this year, expected to bring in a daily average of $100MM in deposits.

2. Creating web services for one of our clients to centralize their fund wiring process through us. Expected to bring in about $500K+ annually.

3. Conversion from our current wire services provider which promtped a number of upgrades for us, including our bank accounting system, implementation of clustered message queueing and Java application servers, application-layer failover and God knows what else before we're done.

Infrastructure:

1. Security upgrades, including a redesign of our AD schema, web applications, web content filtering/blocking and hardening all of our IIS servers. Yes, that whole download.ject thing scared the bejeezus out of me.

2. Management upgrades, including the purchase software to better manage our virtual servers, better patch management, application management, firmware rollouts and provisioning.

3. Upgrade of our CRM software to the latest and greatest. We're only five versions behind so the move is a little painful and is taking longer than expected. The payoff is a more flexible system to incorporate more advanced business processes.

Some of these projects have been very, very slow to build up to completion but they're finally starting to come together. Stay tuned!

New Toy

2004-07-27T21:42:00.000-07:00

My Blue Coat ProxySG 400 came in. I've been waiting for this for a couple of weeks now. It's a proxy server that doesn't only monitor web traffic but can also filter it, strip controls out of it and generally keep bad things from coming back through the network. The vendor doesn't think we can configure it ourselves. Hah!

Fortress in the Making

2004-07-21T21:03:00.000-07:00

How does a company keep things like the latest Internet Explorer vulnerabilities from trashing its network and possibly sending off client data, thus invoking penalties from California's information privacy acts, a scenario that gives me the heebeejeebees?

Well, I'm not sure. But I do have a plan. After scaring the bejeezus out of senior management last week, I propose a plan to help reduce our window of vulnerability (nice pun, that) thanks to IE. My suggestions:

1. Buy a content security management device such as Blue Coat's ProxySG and strip all unnecessary ActiveX and Javascript. Be absolutely ruthless with the content management when possible.

2. Implement client firewalls such as Symantec offers. While this won't prevent infection, it can slow the spread of worms and Trojans by shutting off ports they might want to use. For example, when Welchia got in our network it spread quickly by using pings to find other potential hosts. Stop pings and several worms won't spread. Stop the gratuitous use of Netbios over TCP/IP and you cut down even more. Rinse & repeat -- by eliminating wide-open ports on every machine I can limit the damage done.

3. Maybe it's time to replace Internet Explorer. I use Firefox at home and I think it's great. Maybe we need to replace IE for Internet use, keeping it available for some of the intranet servers we have. My understanding is that I can use a CSM appliance to do that.

The possibility of seeing client data sent to some server by a keystroke-logging Trojan horse gives me cold sweats. If it wasn't for websites crucial to our business that use DHTML extensions that only IE can read and ActiveX controls that only IE will run, I would have ripped out IE yesterday.

More Vendor Management

2004-07-19T20:47:00.000-07:00

The creation of the competitive bid has to be the best negotiating tool ever. For the last two months I have been working two vendors regarding the replacement of a single color copier. Small potatoes, you say? Nay, I say! For this damnable device was costing us about $1900 per month thanks to the foolishness of an ex-executive assistant. After pitting two major copier company reps against each other, I was able to bargain the price down to half of that. Naturally, it didn't start out that way. The existing vendor offered me "lower monthly payments," reducing my costs by $200 per month. How sweet. The competing vendor (who had won my black & white copier contracts about two years ago) showed me where the residual was hidden in that payment and offered me a lower base & service fees per copy. Then it got exciting. Offers got very aggressive. By the time I was ready to call it quits, I was prepared to go with the existing copier vendor. I tried to sweeten their deal by having them take back a black & white copier I didn't need, but no sale. I was a little miffed at that, but they still had the best price close to the deadline.

Time to make a final decision. I made the call to the competing sales rep and started explaining the bad news. His boss was in the room and the conversation went very quickly. He shouted out, "How much are we off by?" I threw out the figure with a little padding. He said, "If I give it to you as a service credit, will you go with us?" "Absolutely," I said. The sales rep was in my office in an hour.

Why did the deal get done so quickly? Because I knew the numbers and knew the other vendor couldn't give me this deal, which was effectively a year of service for free. Since we've been using the competing vendor in other situations, I knew their quality was good and they would give us great service, even if it meant damn near slitting their own throats. The best news, of course, was that we got a pretty reasonable price for what we needed. I had good fun with this one.

The moral of the story? All negotiating books tell you that information is power. So is the competitive bid and knowing what you need. One more thing -- don't be afraid to challenge a vendor you have a good relationship with every once in a while to make sure you aren't getting overcharged. My old boss used to tell me that a vendor relationship gets too cozy after a couple years and shaking it up a bit is a good thing. I agree completely. The bottom line is still the bottom line.

Sometimes I Get Busy

2004-07-11T22:20:00.000-07:00

The last two weeks have been very, very busy during the weekday and mostly very, very mundane on the weekends. Lots of projects underway, including some interesting contract negotiations. Good stuff, coming soon!

Are We Making Progress Yet?

2004-06-30T16:35:00.000-07:00

I think the most frustrating part of my job is when everyone is working diligently on their tasks and I want to get more accomplished. Without, naturally, hiring more people to do it or working my staff to death.

There is a project we've hired out because we just can't make the deadline -- creating a standalone application for our parent company to handle some 1031 transaction accounting. It's going to be 60 days late. Ugh.

Today's adventures included one hour-long meeting to discuss the alarm vendor relationship and who owns it (three separate departments plus another vendor...too many fingers in the pie) and another hour-long meeting

The new IE vulnerability assures me that we'll have another distraction from actually moving the business forward. I wonder if we'll get hit before we put in some kind of web filtering appliance (maybe Blue Coat?) to stave off this latest security threat. I was hoping we could use Mozilla's Firefox, but it appears to break too many websites used for business purposes, including our payroll system.

Some good news. Over the past six months we have:

1. Successfully integrated our sister investment company.
2. Created the infrastructure for banking to create their own statements.
3. Set up the means to process tax transactions from a sister 1031 exchange division.
4. Updated our backbone switching and routing systems, increasing their capacity by a factor of ten.
5. Set up daily banking reports to use our imaging web presentment solution, including converting archived spool files.
6. Vastly improved the quality control of our imaging processes.
7. Updated an internally-written check reconciliation fraud prevention app.
8. Deployed Blackberry PDA's to select executives.
9. Installed backup lines at our hot site for a banking and trust accounting vendor.
10. Upgraded our storage infrastructure.
11. Upgraded our fax servers.
12. Upgraded our performance review software.
13. Installed a leased line between our Glendale office and the home office at the request of executive management.
14. Created a new report generation and viewing process.
15. Selected the middleware component for future application development.

Too much of this is infrastructure, though. Not enough direct ROI. We have several projects, some overdue, that need to be implemented in the next six months or we'll miss out on potential revenue. Of the twenty initiatives I have left for this year, only seven add to revenue while three involve streamlining operational needs and reducing expenses. Of those seven, if this department can get the Big Five completed (1031 transaction accounting, savings automation, wire service provider conversion, positive pay upgrade and wire initiation via web services) we'll be in a position to increase our company revenues by approximately 15-25% annually. No pressure, though. ;)

It's Good to be King

2004-06-30T16:17:00.000-07:00

I'm putting my condo up for sale and getting it prepared is a bit daunting, even if I'm lining up some tried-and-true vendors vetted by our real estate department. One of the advantages of being the IT King is that I can work outside the office using a wireless Internet connection via Verizon Wireless. It's between dialup and cable in terms of speed but it's really convenient when you're waiting at home for half the day and you've canceled your broadband connection.

DR on the (Relative) Cheap

2004-06-30T16:02:00.000-07:00

Last weekend I performed a test of our disaster recovery site. Thursday I spent down at the site preparing it and testing most of the connections. Saturday was the real test, from 9am to 3pm. Most of our servers are virtualized, so I had updated the copies during Thursday's prep work. The data on the servers doesn't change much so it only requires an update perhaps once a month. The really sensitive, daily data gets mirrored using Network Appliance filers (an F940 cluster at our primary data center and two FAS250's at the hot site). Our Exchange information stores, SQL databases, web sites and user files make up that data getting mirrored. Most remote users will access their desktops via Citrix, although we tested locally with volunteers at several branches as well.

Most of the test went well and, naturally, revealed more holes that need to be plugged. We've handled just about all of the must-haves on the relative cheap (at least, cheaper than Sungard Recovery Services has quoted us in the past) and there's still some polishing left. However, it's good to see the DR part of our emergency plans come together so well.

What He Said

2004-06-22T21:51:00.000-07:00

Joel Spolsky writes effusively about the future of Microsoft and its Windows API. The bottom line: Microsoft can no longer monopolize how applications are written because everyone's moving to the web.

My take? Absolutely right. At our company we're using web services to expose access to Microsoft's SQL Server so that we can extract data to PostgreSQL databases and manipulate them in Python and Perl. And so .NET winds up actually cutting off the monopoly power of Microsoft by allowing us access to their server technology without using their "rich clients." Good riddance.

Perks of the Job

2004-06-22T14:19:00.000-07:00

On the eve of signing a very large contract with a vendor, I had the opportunity to be taken to dinner at Pinot Provence here in Orange County.

I started the evening at the bar with some of the Banking Services crew, indulging in a Sauvignon Blanc called l'ete, produced by a Mendocino winery by the name of Saisons des Vins. A nice summer wine, it had a very floral and citrus presence with a medium-bodied finish. Very tasty and more impressive than most Sauvignon Blancs.

Dinner was excellent. I had a lamb rib-eye garnished with peanut potatoes and foie gras. Accompanying dinner was a Revere Cabernet Franc, a delicious red wine with a dry taste and a slightly bitter, acidic finish that went well with the entree. Dessert was a chocolate souffle cake joined by vanilla bean ice cream and chased with a cappuccino. Decadence achieved, I staggered back to the office knowing that at least our hosts have taste in their restaurant choices.

Vendor Management (redux)

2004-06-21T21:49:00.000-07:00

I was forwarded an e-mail from The Girlfriend that described the cutthroat software industry. Remember this well -- if you have a large project and you competitively bid it with a well-defined set of features, you can definitely achieve substantial savings.

In the last two months, two events stand out to me that reinforce the use of competitive bids. I have bid on an upgrade project for our Banking Services unit and reduced the initial price from $270K to $200K. The Banking department manager recently competitively bid out a positive-pay system and managed to reduce the initial price from $250K to $190K. That's $130K over three years in savings to the bottom line. While I don't encourage unethical behavior in bidding, never assume that a vendor has your best interests in mind. Make them sweat. And remember that you are working for the best interests of your stakeholders, no matter how nice your vendors seem.

Pulling My Hair Out

2004-06-21T21:41:00.000-07:00

If you had two lines of business, one netting $3.2 million over the last five months and the other netting ($1.2 million) over the last five months, which one would you feed? What if I told you that the opportunity cost of feeding the loser is an additional $250K per year? Would you continue to sink money into a loser? What if I also included the information that the unprofitable line of business had never been profitable? Would you still continue to allocate it resources or overlook it's cash burn rate?

If you're us, you will. Ack.

Open Source Workflow

2004-06-18T15:11:00.000-07:00

Because I'm characterized at work as a cheap bastard (completely true) and I prefer small, guerilla projects to prove a concept over giant implementations of an unproven technology, I've decided to list some of the Open Source Workflow software available on SourceForge so I remember to take a look at it and see if anything is useful for this organization. Without further ado, the list:

Open Workflow Engine
Java Business Process Management
Java Graphical Process Designer
XFlow Process Management System
u|Engine
Twister

UPDATE:

Ah hell, a little Googling and I found a link to Open Source Workflow Engines Written in Java. Oh, and here's another list.

Business Process Management

2004-06-16T17:29:00.000-07:00

So BPM is the Big Thing, according to all the e-mail newsletters I got in my inbox today. It was Aligning IT with Business yesterday and it may be Real-Time Enterprises or the Data Center of the Future tomorrow.

What is BPM? It looks like it's supposed to be a mating between reengineering and workflow automation, which had so many syllables it produced a three-word offspring. It also looks expensive; one case study places a $200K software price tag plus 3 full-time employees to maintain it. Now ITE is part of Navistar, a company with a $2.7 billion market cap and about $8 billion in revenue. For them, BPM is cheap and saved them a great deal of moolah. For smaller companies that are, oh, about 1/400 of their size, I suspect that the price of BPM does not scale quite as well. Much like our CRM implementation, which financially did not scale well, I would forsee some BPM software implementation sucking up all of our net income and then some.

Is there a way to do some BPM on the cheap? No, it doesn't look like it. Or maybe I'm wrong. It looks like this topic deserves a little more investigation. Interesting. More on this later.

Adversarial Relationships

2004-06-16T16:52:00.000-07:00

In my last post, Rayne comments that when "the client and consultant view each other not as partners...but as hostile adversaries, it's all over." Absolutely true. So what is it that put me in such an unforgiving, hostile mood towards the consultancy I mentioned in my last post? Well, let me tell you...

It all started back in the day when we bid out our last third-party IT exam that the Feds tell us we need every other year. Another Well-Known Consultancy actually won on price, but then the corporate parent called down* and made sure their Pretty Well-known Consultancy won the bid. I'm too lazy to fact-check this, but rumor has it that there is a bit of a revolving door between the Pretty Well-known Consultancy and our corporate parent. In any case, there was a great sound and fury and suddenly the Pretty Well-known Consultancy matched the best offer and they came in.

The interviews lasted for a week. Of the three people brought in to review my department, two of them didn't know what they were doing and refused to actually review material I provided them because it wasn't in the format they requested. I started to get peeved. When I saw a draft of their results I went from peeved to pissed. I called up the project manager, whom I thought was reasonably savvy, and asked her how her group came up with their recommendations. She said the magic words, "Best Practices." I then explained that we are a small firm with less than 100 people and $12 million in revenue, did these practices make sense? Wouldn't it be better to actually review our sector of the industry and recommend practices that wouldn't increase our relatively healthy IT budget by 50%? She said she'd consider it and then published the exact same results as the draft that I saw initially. I spent about two hours in front of our audit committee challenging their results and getting a favorable response. For what we paid and the time it took, the results were appalling simple-minded and showed a complete lack of appropriate cost-benefit and risk management tradeoffs.

The second time the Pretty Well-known Consultancy came along, it was for our SAS70 exam. Similar group of people, similar results -- best practices that might be appropriate in a $100 million company, probably excellent in a $1 billion company, but ridiculously expensive and counterproductive in a (now) $25 million company. I asked the same question, "Are these the best practices for a company of our size in this sector" and received the same non-response. Once again I spent about two hours in the audit committee explaining to them why there was a divergence between our practice and their recommendations. I applaud myself for the fact that I never used the term "horses**t" even once. I am also beginning to accept that this farce will be played out every year until we reach a size that is appropriate for the recommendations.

Why is it that we aren't acting as partners? Because there is no reason for the consultancy to alter their standard cookie-cutter pattern. They know they have our contract every year. There is no additional incentive to provide us useful information that would require them to actually understand the SMB banking & wealth management sector. Why align with our actual needs if they can just do a quick job and publish a report that, quite frankly, they have a monopoly on?

Corporate ass, er, backscratching at its finest.

* This is what is known as uncorroborated but probably true, considering the source.

Sclerotic Organizations

2004-06-10T14:18:00.000-07:00

Why do companies find themselves moving slowly, unable to innovate, as time goes on? I believe it begins with the laudable goals of standardization and the implementation of change controls. Standardization starts out with the thought that configuring the desktops and servers with a certain allotment of software will streamline operations and drive down costs, which it does, until it hits diminishing returns for the sake of the bottom line (Look! We have one desktop support person per 1000 users!) Standardization then works its way into application development, where the goal of finding the best solution becomes instead "find the best solution with the following platform and language constraints." It's a worthy goal that begins to transmogrify as it is applied, reductio ad absurdum, to every interface, every script, every problem. Then the strangulation of innovation begins.

The second hand in this death-grip is change controls. Change controls exist to reduce risk. The spectrum of change control is broad, from the idea that procedural change controls will be sufficient ("don't bring down a production system or you'll be fired") to the creation of review boards that must sign off on production changes (a la JP Morgan Chase, which evidently has three such review boards for just their production network). In addition, we recently received advice from a Pretty Well-known Consultancy firm that suggested we should record every change made to production, whether systems or appdev, and be able to produce a report that linked every change to every approved request. I want you, the reader, to understand -- we're an IT department of six people working for a $24 million a year company. This Pretty Well-known Consultancy firm has mistaken us for the Department of Motor Vehicles.

I think that the management of risk has become less of a cost-benefit question than a cover-your-ass proposition. This post-Sarbanes-Oxley environment is turning otherwise thinking, intelligent people from beings capable of rendering their own opinions into the lowest common denominator best-practice zombies so that no one can't point fingers and say, "You're wrong!" or "You gave me bad advice!" Grow up and grow some cajones.

******* Lakers

2004-06-06T21:44:00.000-07:00

You know, the true test of management is getting a team to actually execute. Larry Brown and the Pistons tonight showed that they can. Once again, the Lakers showed that they can underchieve with the worst of them.

But there are three more games to be won...

You Know...

2004-06-06T14:54:00.000-07:00

...I had this big long post about how I do my budgeting (expense as run-rate, capital as zero-based) and set it up to give me some flexibility in case I have some unfunded project drop in my lap, like upgrading the Banking Services department's production machine and setting up a high-availability system at our hot site. And, right now, I just don't care.

Right now I am sitting at home listening to blues on KKJZ, drinking Old Guardian, another masterpiece beer brought to you by the folks at Stone Brewing Company and cleaning this place up in anticipation of the Lakers game where I expect to see the Lakers destroy another worthless Eastern Conference team.

As much as my job and industry attempt to consume me, I've still got a life.

Speaking of Writing Memos...

2004-05-25T21:47:00.000-07:00

Thinking back to the memo I wrote, I remember reading Chapter 12, Interacting with the Business in the EGIT. It is a short chapter and yet one of the most important ones in the book -- how to maintain effective relationships with the business. Boiling it down, it's about establishing good communications and relationships.

I like the example e-mail message that rambles on. I find it equal parts funny and scary because one of my staff talks like that...yikes. The counterexample is a good one and I will probably use it in my operations manual once it's done. To summarize:

1. Be clear.
2. Bottom-line your message.
3. Remember the basics of grammer, punctuation and formatting.

I'd add a fourth item -- remembering the ambiguity of e-mail, it might be better to not send it and make a phone call instead.

Establishing relationships over informal meetings is important as well -- I do semi-regular walks with the HR director, once a month lunches with the Operations and Real Estate department heads and work fairly well with the head of Banking. I need to spend more time schmoozing the regional VP's and maybe work better with the finance people so I've got allies surrounding my only major obstacle on senior management (and only an obstacle because of factional fighting. Sigh). Politics is a part of my job I do not spend enough time working on. My mistake...and one I hope to correct with the staff add I mentioned prior.

How to Get a New Employee Without Really Trying

2004-05-25T21:30:00.000-07:00

After some dithering, I've decided to move forward in my pursuit of another staff member. I was reminded of some possible objective measurements by John Bashab, one of the authors of The Executive's Guide to Information Technology (EGIT), an excellent reference book for all things IT management-related. One is mentioned in Chapter 13 regarding budgeting -- how to assign labor in the budgeting process. One of the issues flagged by the authors is that an increase in budgets usually signal that more labor is required, hence more staffing may be necessary. In our case, our budget has increase about 25% year over year due to expanding business opportunities and an unexpected merger.

I'm not one to leap to spend money on more staff and build and empire of people. As a matter of fact, I want a lean and mean organization. I do not want an IT staff that sits around doing R&D all day because it doesn't have enough to do. On the other hand, I am aware that our "special projects" come in bunches and that taking the time to share knowledge between staff members and with other departments is valuable. We're reaching a point where that is taking away from other duties, both operational and special. Hence, I put forth the request to get another staff member. I intend to hire in another bottom-of-the-rung person to fill the desktop support position and promote the existing person into a systems engineer position because he has the talent and drive to succeed with that. After taking an inventory of the number of servers we have (read: operational needs), the number of business initiatives on the table and the number of IT initiatives I'd like to push, it just made sense to write the memo and justify another person.

Another Damn Manual

2004-05-21T15:53:00.000-07:00

More of my staff's time continues to be taken up by regular and repeatable processes, i.e., backups, audits, patches, etc. I've decided to have them identify those processes and commit them to an operations manual in order to standardize what they're doing and how they're doing it, plus build some more exception reporting into the mix. After putting this together, the next step will be figuring out what can be automated and what can be rotated. Automate any checks or reconciliation, rotate the manual tasks so that nobody gets complacent and everyone knows the small processes. Everyone wants to ride the horses but no one wants to clean the stable.

Product Research

2004-05-19T22:21:00.000-07:00

I spend a very small part of my time meeting with new vendors and getting new ideas. One of my recent meetings yielded some interesting information, including introducing me to two of their partner companies that sounded promising in fulfilling some of our goals.

The first company, Blue Coat Systems, makes an appliance that works as a content filter/cache/IM control/web virus scanner/bandwidth controller. I was looking at WebSense as a solution to control and report on web surfing, but this has my attention. Blue Coat used to be called CacheFlow starting in 1996 and had a pretty good IPO if I remember right. So they've been around for a while.

The second company is call PlateSpin. They have two major products, one of which converts physical machines to virtual machines (and virtual to virtual for the different flavors of VMware products) and another that manages/provisions/deploys physical and virtual servers. Given our dependence on virtual systems, this one also has my attention. They are a Canadian company and appear to be a fairly recent startup.

What Is My [Big] Problem?

2004-05-18T17:58:00.000-07:00

In a column for Infoworld called CTO Connection, Chad Dickerson asks the question, "What's your Big Problem?" The Big Problem is the strategic issue the company has that you're stumbling towards solving.

That got me thinking...what's my Big Problem? Do I have one? Hmmm. From a business perspective, there's "How can we offer more banking services?" and "How can we retain and grow our wealth management business?" The first leads to thoughts of a flexible, highly-available and easily administered infrastructure. Or something like that. The second thought is more along the lines of integrating existing sources of data and turning it into information that can be used to retain and sell new prospects.

From an IT perspective, all of these things are possible by spending enough money. Here's the trick: how do I reduce the daily operational costs and staff time to free up the people to do the projects that bring in the real money? That's the Big Problem I whittle away at day after day...

Taking Over a Low-Morale Workforce

2004-05-18T14:31:00.000-07:00

I came across an article on ComputerWeekly.com via a bunch of semi-regular e-mails I receive through an outfit called IT Business Edge. Usually I barely skim this stuff, but today I'm inbetween some tasks while updating our hot site. Anything to do with people management gets my attention, especially issues with staff.

The article itself is from several viewpoints discussing how to turn around an unhappy IT department. My favorite answer (only because it must be satire) was the one from a member of the CW500 Club who suggested firing 25% of the staff until morale improves. Chuckle. That one had me going until I realized it must have been filler because the CW500 Club is a service offering from ComputerWeekly.

Seriously, one of the answers got my attention -- is it the department or the culture? Once that's been defined, the department head can figure out what's going on. Good luck changing the culture. Therein lies a quick exit out the door. Changing the department is easier, and some of article's suggestions were sound. Most likely it's because the department did not know what its priorities and expectations were. One of the greatest challenges I have is mapping out clear priorities and expectations. Not everyone follows them, but at least they're there. ;)

Tired and Uninspired

2004-05-17T17:37:00.000-07:00

I woke up dead today and it hasn't gotten any better since. Even a trip to the local Starbucks for a trip mochafrappacremedechino or whatever failed to jumpstart me. Bleh. Today I mostly stayed out of everyone's hair, except for a meeting with another department head and the company that's programming an application for us (i.e., we're outsourcing a project to India to keep headcount stable). Snooze. My only real involvement with the project is to prevent creeping featuritis and keep it on time.

Where Are We Falling Short?

2004-05-13T09:24:00.000-07:00

Generally speaking, the business initiatives we are pursuing are getting accomplished. In fact, they appear to be moving faster than before, perhaps because we've really held the line on user support by continuing to reduce help desk calls even with the addition of staff. We're now down to part-time desktop support for 110 end-users on a normal day. Most support calls over 15 minutes in length are handled by Level II support, aka the system or network engineers.

As good as the implementation of new initiatives is coming along, there are areas in which we are not paying enough attention:

1. Security.
2. Standardization.
3. Operational excellence (i.e., making sure the little things are getting done right).
4. Documentation.
5. Change management control.
6. Deeper understanding of new systems rather than Just-In-Time troubleshooting.

So is this accomplished by the addition of more staff or a slowing of existing projects? Or am I not applying their skills effectively at this time so that they are not achieving their potential?

Time to Add Staff?

2004-05-12T21:18:00.000-07:00

I've been considering lately whether I have the right amount of staff and the right mix of skills. While reading The Executive's Guide to Information Technology, the authors had made the case that IT consisted of the Application Development and Operations functions. I am thinking that there is a need for another function -- Special Projects.

The IT operational function should be concerned with sustainable and standardized processes. Backups, security audits, monitoring and managing server uptime, upgrades, reporting, patch deployment and the help desk would all fall under this category.

Special Projects is about tackling new challenges -- implementing a new banking function, planning the upgrade and user acceptance of a new web-based CRM or evaluating the existing disaster recovery plans and recommending changes.

Looking at the functions, I need to ask myself if I can rotate these among existing staff in order to challenge and educate them, or if I need to concentrate that expertise. And if I rotate, do I have enough staff to execute these projects in a timely manner?

My current staff (including one contractor) looks like this:

2.5 software developers
1.0 desktop support
1.0 system engineer
1.0 network engineer
0.5 project manager

Is that enough? With 110 end-users, the answer would appear to be "of course!" Really? What if I break down the metrics this way:

110 end users
5 branch offices
80 servers (yes, 80)
4 operating system platforms (Windows, Linux, OS/400 and OpenVMS)
99.99% uptime requirement for banking-related systems

Now do I have enough staff? Maybe not.

This thought to be continued...

More Blade PC's

2004-05-12T21:03:00.000-07:00

What is going on with the world? More PC blades? Yikes! Once again, I question the need to combine the worst aspects of thin clients, blade servers and provisioning software. I understand why a company would want the control of thin clients. I can certainly see the necessity of using provisioning software to restore systems quickly. What I can't fathom is why you'd stick a bunch of PC's in a rack, connect end-users to them via a lobotomized thin client station and still need to purchase provisioning software to roll out software images to all of them. Ridiculous.

There Is No Greater Reward

2004-05-12T18:12:00.000-07:00

It's been two days since I've been back. Gone for a week and, after slogging through a thousand e-mails, I've been waiting for a shoe or two to drop. Nothing. No cries of pain and/or agony. Nothing went wrong at all. Everyone did their job, watched out for potential problems and basically kept it all going. There is no greater reward for a manager than knowing the department is in good hands while he's away.

The Mobile Workforce is a Ticking Time Bomb...

2004-05-12T10:18:00.000-07:00

..for regulated industries at least. Upon what do I base my conclusion? Two thoughts:

1. We are currently preparing for a visit from the SEC to examine the company after merging with our sister investment division. They are interested in reviewing, among other items, marketing material created and used by the salespeople. Well, most of the salespeople in that division used laptops and stored their data locally. When the laptops came back, would you be surprised to find that everything was wiped out or not working?

2. Some salespeople are allowed to keep their prospects in their own contact managers (on laptops) rather than placed into our CRM package. I've said it before -- it will bite us in the ass before too long, especially if the regulators come knocking or the salespeople start calling prospects that we don't know about because they aren't in our system.

Decentralized control in a regulated industry is idiotic.

Back From Vacation

2004-05-10T19:38:00.000-07:00

I just got back from a week-long vacation. Starting at the Port of Los Angeles, The Girlfriend and I hopped aboard a five-day trip up the coast she booked through Vacations To Go, stopping in Seattle, Victoria and Vancouver. Once we arrived in Vancouver we stayed two more days in a big pink B&B known as the West End Guest House, visiting with friends and burning up lots of CompactFlash cards taking pictures. Look forward to new posts soon!

Business Makes You Dumb

2004-05-02T13:35:00.000-07:00

At least as far as your vocabulary, that is. Too many memos and e-mails, denatured and devoid of spirit, have reduced my vocabulary into a mere shell of mono- and duosyllabic words. My girlfriend and I were talking about which kind of Mother's Day cards I should pick up this morning. She mentioned that her mom likes the type that are overly sentimental and couldn't think of the exact word. I started racking my brain and couldn't think of the word that wanted to rise to my lips, a word I obviously haven't used in quite some time. I walked three city blocks in the hot sun to get my hair cut. Couldn't remember. Walked to the grocery store to pick up water and the aforementioned cards. Couldn't remember. Finally, after the 90 degree sun had cooked my brain for a bit (perhaps the haircut provided better access?) it finally came to me: maudlin. So it took me about two hours to remember a word I probably learned in junior high. Pitiful.

Marketing IT

2004-05-01T12:16:00.000-07:00

Another article in CIO.com addresses how to present IT to the rest of the business in order to get what you want -- a better department reputation, more visibility in the company and easier access to purse strings.

At my company this is not as much of an issue for me as I report to a single executive rather than a group. That simplifies my message and makes it easier to couch projects in terms that provoke a positive reaction. I admit that it would be draining at this point in time to fight battles with the rest of senior management as I'm still about a 25-50% hands-on guy and projects would slip if I had to play diplomat and politician full-time. I hope to get to that role at some point -- right now just isn't the time.

How to Run IT Like a Business

2004-05-01T12:04:00.000-07:00

An article in CIO.com discusses (yet again) the topic of running IT like a business. To summarize:

1. Standardization.
2. Process and project management.
3. Financial reporting.
4. Communication to senior management.

The article is still a little fuzzy on how these IT departments deliver value except for the Intel example where they tag a dollar value on just about everything. It does make me wonder how they justify infrastructure upgrades and the like. Are server upgrades given ROI's? If so, on what basis? No answer. Some other categories of measurement:

* Internal customer satisfaction scores
* IT leader performance reviews
* Executive committee reviews
* Industry benchmarks
* SLA performance

In my company, I am pretty confident that I meet customer satisfaction. Since I'm the only IT leader, I think my performance is OK. We have no formal SLA's beyond uptime requirements. And industry benchmarks are a bit misleading, as there are few wealth management/commercial bank (sans lending) combinations quite like ours. Do we really compare to the aggregate?

A recent Optimize article highlighted two financial institutions in a buy vs. build discussion. How do you compare these two companies? Ratio of IT personnel to total employees? IT spending as a percentage of revenue? Return on each project? Depending on how you slice it, a department may be doing well, so-so or poorly. The numbers selected determine the agenda of the person doing the slicing and dicing.

I Had No Idea...

2004-04-29T20:58:00.000-07:00

Thanks to spending time doing nothing but reading Joel on Software, I have discovered that everyone at Microsoft is getting paid to write a blog. I'm in the wrong business. Somehow, I must convince my boss that I need to blog all day and collect the dough every couple of weeks. Soon I will be like Dogbert, with a consultant job for which I telecommute and receive pay via direct deposit.

A Question Every Manager Should Ask

2004-04-29T20:47:00.000-07:00

Pete Kruckenberg asks why people don't emphasize the potential strengths of others instead of trying to fix their weaknesses. I understand the desire to make everyone stronger and help them to be more of what they can be. In the management world, where Pete aims his last comment, let me give my perspective. Let's say an employee is great with people but lousy with detail. This hypothetical employee wants to be in an IT department. A lack of attention to detail is a killer if this person wants to be anything but a help desk support specialist or a trainer. That's a weakness that needs to be "fixed" before someone can advance. Strengths need some encouragement and direction to grow. Weakness needs more than that -- it requires discipline and effort to overcome. By focusing on improving weakness, a manager can round out someone with potential.

JBoss

2004-04-29T20:17:00.000-07:00

I had a call with JBoss today regarding their suite of products. We're mostly interested in their middleware (JBossMQ and discussed that to some extent. Primarily it was their attempt to be "warm and fuzzy" although the sales guy and support engineer need to learn one basic rule -- do not talk over the potential client. Ever. If we hadn't already pretty much made up our minds about this being the right product to use, they would not have swayed me. Sheesh.

When I get back from vacation they'll be discussing with us how JBossMQ clusters. We're thinking about using PostgreSQL as the message repository backend. Specifically, Command Prompt, Inc. offers a commercial version that supports replication. I haven't called them yet today, but there's still time in the vendorfestathon.

More Vendor Stuff

2004-04-29T19:59:00.000-07:00

I get swamped by e-mails and junk mail from vendors, so I'm continuing to share the love. I followed a vendor link from Active Reasoning and downloaded their whitepaper, "Change Management - The Other Half of the Story". Although a small company, we often find ourselves needing to explain why we don't have a formal change management process in place that tracks every change to our systems. These folks appeared to propose a solution in their whitepaper, so I read on.

Like most whitepapers, it's pretty thin on detail. Apparently the firm will custom-build (btw, that's vendor-speak for "really expensive") a solution to track, approve and log changes made to any system specified in the scope of the engagement. Heh. Did I say, "really expensive?" I'm thinking freakishly expensive at this point.

Don't get me wrong. I'm not knocking the concept at all. I would be very happy with something in place that would make it easy to tie back every change to a request, log when it occurred and who did it and produced reports that would get regulators and examiners off my back. However, I'm not interesting in a bespoke application that will only cost me a few appendages and require changes itself as time and objectives move on.

Secure E-mail Delivery

2004-04-29T19:52:00.000-07:00

I was in a vendor meeting today to discuss how to send out wire notifications securely to our clients. You see, most of our commercial banking clients want to receive a notification when we have processed a wire on their account(s). The e-mail would contain the banking accounts, routing numbers and other assorted info. The goal is to get the wire confirmation into the hands of an escrow officer as soon as possible so escrow can close and our client, the escrow company, get paid. However, that's some pretty sensitive client information in that e-mail, so sending it plain text is mighty dumb. What to do?

Which gets us back to the vendor meeting. I discussed the above problem with a company called PostX which has a secure e-mail solution that sounds interesting. Here's what it will do:

1. Application generates an e-mail.
2. PostX server scans outgoing e-mail, recognizes it according to a business rule and encrypts it by changing the actual contents into an HTML attachment using a digital signature.
3. Client receives the e-mail and, if opening it for the first time, registers on the PostX key server that we would host.
4. Once registered and their key downloaded, the client can open any encrypted e-mails without re-registering.
5. Client reads e-mail and can save it for re-reading later.

Kind of nifty and doesn't require any additional software installed on the client desktop, which is an important factor as we evaluate solutions for this. More later.

IBM Joins the Virtual Server Game

2004-04-28T15:36:00.000-07:00

IBM has decided to join the virtual server fun with its own "virtualization engine", essentially doing what VMWare is doing on a Power 5 processor. Although the product is not officially announced on IBM's site, it appears it will be released at the same time the Power 5 processor hits the market. Along with it will be new management and provisioning software. Now if IBM can stuff 2-4 Power 5 processors in a blade and get any kind of performance out of it, I'll be able to put just about my entire data center into 7U. I *like* that.

Guy Reviewing Referral Logs

2004-04-27T21:15:00.000-07:00

So I had some interesting visitors today. I thought the highlight of my night would be watching "VH1 Goes Inside South Park" but it turns out I had a good time looking through my referral logs for the last couple of days. Let's see:

europarl.eu.int -- yes, not only have I had a visit from eop.gov (Executive Office of the President of the United States), I now have a visit from the European Parliament. Am I cool or what? It was off of a search for BizTalk and JBoss. I'll summarize in case you come back: BizTalk is damned expensive and complicated, while JBoss appears to be inexpensive and one of my developers got it up and running in three weeks. We started our first message passing today with the basic install, knowing nothing about Enterprise Java Beans. Highly recommended.

nih.gov -- looking for Hermes and OpenJMS. I don't believe either are ready for prime-time from what I've read. I wouldn't commit my financial transactions over them.

I may not have any useful info but at least I rank high on search engines.

Blade PC's?

2004-04-26T18:49:00.000-07:00

HP is shipping blade PC's, which link a thin client to a blade running a Transmeta processor. Huh? Who thought this was a good idea? I understand Citrix and server-based computing. I understand blade servers. I understand virtualized servers. But what the hell is this? It's PC Anywhere by hardware and makes no sense to me. What marketing genius thought this up?

Marketing Weenie Analyst: "The industry is migrating towards the centralization of information technology resource management, building a resilient, distributed, utility grid computing infrastructure. This change wave has catalyzed the storage and server market segments, yet left untouched the enormous potential of leveraging the centralization paradigm to the desktop platform. With this in mind, I suggest we find synergy between the consolidation of data centers using blade servers and the TCO reduction of thin clients. To wit, the blade PC, at $1399 each."

Marketing Manager: "You are a f*****g genius. Let's roll!"

And so it was in the land that the worst of blade servers and thin client computing were synergized and brought forth. And lo, let it be said that the PCjr has company in the Desktop Hall of Shame, for this shalt be a real stinker.