I got up yesterday morning a few minutes earlier than normal, and started at work around 8:30. Lesly and I drove to Geelong and joined Richard. We worked as normal until around 5pm when we then went out the back to the server room and began our “real” work day.

Over the course of the next five hours we decommissioned two servers, changed the role on another, comissioned two new servers into production (one replacing one of the old ones and taking the other’s role, and a brand new one), as well as implement a new Division-wide software package, backup schedule and regime and new web server.

Completely knackered and tired, Lesly drove us back to Melbourne — we picked up Justine on the way home from her big night out — and we arrived back at around 12:00. The day was almost over, but I had to check on yet another server that appeared to have died while we were in Geelong. Luckily it didn’t take long to revive it; it had actually been shut down by someone from Geelong so there was not really a problem.

So, my work day yesterday lasted around sixteen or so hours. Definitely one of the longest ones I’ve worked in ages. We did manage to do quite a lot in five of those hours, and they were the result of a lot of planning beforehand which made things a hell of a lot easier.

Just before leaving work I noticed that and LP of Never Mind the Bollocks, Here’s the Sex Pistols was left on the table (someone recently brought in a large pile of LPs for people to take if they wished), so I grabbed it (not that I have anything on which to play it!). Shortly after we arrived home, I was in bed and asleep.

So, last week’s work was just an appetiser; a taste if you will. This whole week, I have been nibbling on delicious Primary Domain Controllers, savoury Backup Domain Controllers, crunchy Domain Name Servers and chewy dcpromo executables, just to name a few. As well, I have bitten down hard on an overripe grape, an NT4 upgrade, which turns particularly sour when upgrading to 2000 in BDC emulation mode.

What the hell am I on about? Well, at work we’re attempting to upgrade our NT4 domain to an Active Directory domain. This is fine, except that we aren’t the top-level in our domain, we’re just one tree in the forest. It gets difficult where the top-level won’t play, since the they’re not a Windows-based domain.

We’ve tried to do this upgrade before — two years ago — and it was not successful. The approach we’re using this time is slightly different, so perhaps it will work this time around. Who knows?

I’ve already installed NT4 probably half a dozen times this week (due to what seems to be faulty hardware), as well as playing with 2000 and having a lot of trouble joining the upgraded NT4-2000 server as a BDC using dcpromo. Luckily, this was a test domain we implemented, and all seems to have gone swimmingly thus far. It looks like the chewy, creamy centre of a successful AD implementation is within our reach; all we need to do is bite through the bitter outer coating of interoperability.

Since no one will probably read this before we start, I’ll have to wish myself luck.

Just got back from work, where Lesly and I have just spent around 6 hours fiddling with a rack.

We had to install a Dell PowerEdge 2850 server and a Dell PowerVault 114T tape unit into the rack. Easy enough, right?

Unfortunately, due to the existing equipment in the rack, we also had to buy new, longer shelves, and move the vertical rails all the way to the back. This meant that we had to remove everything and setup the entire thing from scratch. Least we managed to make it much neater in there, even with the addition of another few cables.

We also installed an el cheapo non rack-mounted KVM switch which didn’t seem to like the trackball on our keyboard. Ah well, an optical mouse is now in there too.

What an enjoyable “long” weekend I’m having. A whole day off!

So, my understanding of the website redevelopment myself and Richard have been “asked” to do was a little off.

I got into work earlier this morning, and another couple of emails had been sent around regarding the redevelopment. I also re-read the original email sent out. There was this one particular phrase which I didn’t fully read:

the majority [of the project] will be done on overtime

Umm, hang on. I’m happy to work a few hours overtime if we’re unable to get the job done during regular office hours, but the majority? I don’t particularly like being “forced” to work outside of office hours, no matter how good the pay will be.

Update (09:35): I might just clarify here. If there is only 5 or 10 hours work (1-2 hours per week), then I’m not particularly bothered. If the project ends up being 50 extra hours, I don’t feel like working another 10 hours per week “under the pump”.

Also, the redevelopment is — as far as I can tell — quite easy. The mockups I mentioned yesterday were time-consuming to make, but once you have made the “master” one, the rest is quite a simple task.

Assuming Richard and I take ~550 pages apiece, and it takes 10 minutes per pages, we are talking over 90 hours each. In one month. Over and above our normal work hours. Even so, not all the design stuff has been decided.

I am still working on a rather large project at work (detailed here) and am making pretty good progress on said project. For those who care about technicalities, a Visual Basic app talks to an ASP server using XML. The ASP server then discusses with a Microsoft SQL server the possibility of storing and retreiving records. There’s more to it than that, but that’s the basic gist of it.

At the start of last week, once I returned from ITIL Training, I was asked (asked is such a nice word for what really happened) to ensure that the project was ready for testing at 9am sharp on Monday 24th May. Essentially this means it needs to be ready on Friday 21st May (unless I feel like working all that weekend in the office, which I don’t).

Now, based on the progress I am currently making, I think the project (which entails the stuff I mentioned before, but from the user’s points of view, 2 Visual Basic apps and a website) should be ready for testing. They want to test it for a week (originally they allotted only one day of testing; ha!) which I still think is too short but will have to do.

This morning, I was again asked to ensure that our entire Division’s website (it’s not my design, let’s just make that clear) reflects the new university-wide templates by the end of June. I just checked, and on the website, there are over 1,100 files with either .php or .htm(l) as their extension. This means they probably all need to be changed. I have already created a number of different samples we could use but they haven’t even decided what they want it all to look like. I’ll be getting help from the new guy — Richard — as well so that will help a fair bit. Still, we have over 500 pages each to change in a month. Now, I had been under the impression since around October that I we wouldn’t have to be responsible for this, since we already have enough bloody work to do. The powers that be have now decided that not only do they want us to do it, but they want it done now, as opposed to October/November when it had originally scheduled to be rolled-out. This is in addition to everything else we need to do.

It was nice of my boss — Glenn — to offer me to take as much overtime as I need to get both of these projects done but I would really prefer to have a light workload than have authority to take any overtime I want. At least count, there are another 23 IT-related projects people want us to complete by the end of the year; there are two of us (myself and Richard) providing support for over 120 users across five sites at the same time.

Hence the “Hooo boy” title for this post. Eep.

At work this evening, Glenn and I had to perform a few bits and pieces of “server maintenance”. We waited until most people had left, turned off the BDC and what followed was an almost Kafka-esque comedy of errors.

Part of the “maintenance” we wanted to perform was installing a new hard disk into a server. This server, a Dell PowerEdge something-or-other cost more than $10,000 three years ago. The server is all-SCSI, two CPUs, three power supplies (for redundancy) and about eight SCSI hard disks. Since all the hard disk bays are taken, we thought it might be fun to try and see what happened when we installed an IDE hard disk (just for backups, not really a production-use disk).

Well, first of all, it didn’t fit. Dell has devised a form factor which requires you to buy adapters from them if you actually want to add any extra, special hardware. Not to be beaten at this stage of the game, we poked and prodded, added a hard disk mounting bracket (which was too small, but could be made to sit securely in the space) and started installing the IDE controller card. This small setback would have been too simple. Next, the IDE cable wouldn’t reach, since this server is so long (deep). We managed to find an older one which did reach, but it severly reduces the hard disk’s speed from ATA-133 to ATA-33. Oh well, let’s try it anyway!

Now, the server won’t boot! What?! The server was looking for boot information on the (unformatted) IDE hard disk. Well, let’s tell it to search for the SCSI first. Nice and easy. Except that this doesn’t work.

We decide to abandon this since we’ve already spent over 1.5-2 hours trying to get this up and running. Mental note for the future: “It ain’t gonna work!” Next thing is to put a new server into the server rack, and get it up and running. No sweat, I’ve already installed this server, attached it to the domain and it’s got all it needs installed.

Hey, wouldn’t it be great if this new server could use the hard disk, since the older server won’t? So, I connect it. The new server does detect it, but since it’s older hardware doesn’t know what to do. Oh great. I limit the hard disk from 120GB to 32GB which allows the BIOS to find it, and boot.

We’re on our way, great! Now, since the new server was given to some people to install software on has returned, it won’t behave on the domain as it should. It can’t see other computers outside of our domain. Ugh, great…

So, long story made a little shorter, we muck around for another hour trying to get this server running, but it won’t behave. Tired, and a bit flustered we decide that “near enough is good enough” and both leave for the night.

The BDC is at least working, and the new server sort of works… Near enough.

Wow, it’s Easter tomorrow. That doesn’t really mean a damn thing to me though, as I have to work both Friday and Monday. This isn’t necessarily a bad thing though, as for working those days I’ll end up getting an extra 3 days pay for just working. That’s not too bad I guess. I’d actually rather have the time off but if I have to work then at least I get paid for it (not well enough, but since when do I ever get paid enough).