The Grumpy IT Admin

Well no upgrades happening this weekend due to compatibilty issues and trying to schedule down time for servers. Due to a lack of foresight by the powers that be, we do not have a scheduled period where we can take servers down for maintenance, to my mind this needs to be set in stone in the following way "between x time and y time on this day you need to expect your server to be restarted", this gives us a windows when we don't need to worry about whether data is be imported or worked on and there is no other real activity of note being carried out. Just because it is there doesn't mean we will use it, but it is nice to have the opportunity without jumping through hoops trying to get permission from a bunch of system owners to update drivers and reboot a server. This time it was both VMware and EqualLogics updates that needed to reboot servers. Some of our servers have direct attached iSCSI drives on the EqualLogic, these run the HIT kit and since version 5.0.2...

Hopefully if plans go well I will be doing a bunch of upgrades this weekend. Firstly I will be doing the EqualLogic firmware, which, due to some rather buttock clenching moments, I have got different versions on a couple of our systems so I need to get them all in line. I will probably go for 5.0.5 rather than the latest 5.0.7 as it has been out for a while and seems to be relatively safe and as I get older safest seems more bearable than latest. This also means I need to get all the iSCSI connected servers on to the latest version of the HIT kit which is always a worry as most of these are physical servers so you can't just snapshot and then roll back if there is a problem, I will probably pull one of the mirrored system drives before updating to give a quick recovery if it all goes horribly wrong. The other is an update to do is VMware up to 4.1, this is a straight forward thing for the most part, however it is made more problematic by the need to update the VMware tools on ...

Had a conference call planned today with a Dell boffin and our external support people to discuss firmware and all stuff EqualLogic. Unfortunately said boffin was OOO but no one thought to mention this until we had been waiting on the phone for 10 minutes and then emailed our Dell contact (who was on paternity leave). Rescheduled for Monday but once again Dell fall over on simple Customer Service. We have about 30 or 40 companies we deal with, either support or development and to be honest there are only three that I feel are actually any good. One is called PDQ which we have hardware support with for our servers and printers, they are efficient and always follow up any deliveries to check they have arrived and for generally leave you with that nice warm glow of feeling wanted, we also have always dealt with the same people which makes a huge difference. Second we have Voyager our networking people, the account mangers are salesmen and can be irritating but they have a diamond in t...

As we rush headlong over to Exchange 2010 (don't get me started) one of the jobs I've had to do was upgrade our Enterprise Vault to version 9.0.2, this also brings with it support for Outlook 2010 at the client end allegedly, but for now we have a 2003 Exchange environment. The first attempt at an upgrade only just managed to get out of the starting blocks before going pear shaped. I patched the server installed SQL reporting and moved the Message Queue to a different location as recommended by the Vault Deployment Scanner and after reboot we had server that was crawling along a wouldn't start the vault services. Our support company were as much use a the snowball in the Sahara. We managed to recover by going back to the just finished backups of the systems drive and all was well again. I did a copy of the server and tested everything again and was unable to replicate the problem so tried again on the live server with out doing any patching or other stuff and all was wel...

There is nothing like doing patching or firmware upgrades to really get those buttocks clenching. After so many that have gone wrong it has become one of those jobs I'd rather go on holiday when they are due and leave to someone else who is younger and more care free and has yet to be scarred. The latest round of EqualLogic patching at one of our sites went a bit tom tit. Upgrading from 4.3.7 to 5.0.4 should have been a fairly simple thing, though previously I did experience a small problem with one of the controllers restarting about 15 minutes after the upgrade which was a bit weird I wasn't expecting any issues. However this went completely bonkers after the upgrade. Lots of errors which were triggered by cloning a replicated drive. Replicating would start to fail, drives would disappear and then reappear, the only solution that Dell could come up with was to move to version 5.0.5. as 5.0.4. This was weird as I had been told previously that I should move to 5.0.4 as it w...

Another one of those weird things that happen to us admins. This time it was our HP C7000 chassis. On and off we have been getting some errors where the Chassis reports some problem then 1 millisecond later all is well again, also we have had VMware reporting loss of redundancy on the iSCSI connections and packet floods logged on the chassis. Whilst I was off enjoy the sunshine this went completely ballistic to the extent I was getting emails at the rate of 1 every 1 to 10 minutes. I returned to work to find 4000+ emails, not a bad score for 4 days off. A quick Google and the best advice I could find mentioned reboots can make it go away for month but best to do a firmware update, though these mentioned old versions we were on 3.11 and the latest was 3.30 so maybe worth a shot. Given my experiences with EqualLogics (did I mention that another site upgraded their EQ to 5.0.4 and lost all their replication, barely managed to make it through the day with drive intact and had to upda...

I read a few blogs and tech sites that started of as blogs (sometimes before the world discovered blogs) and the one thing that tends to strike me is that whilst the content is often very useful and informative the command of the English language of some bloggers is really quite poor. There are, perhaps, several reasons for this, things like lack of a spell checker in the built in blogging tools on websites, the need to blog there and then, phone texting and street language all spring to mind. Twitter is is prime site for poor grammar and misuse of words with too and to and your and you're seemingly being interchangeable and having no separate meanings. Whilst I can forgive the immediacy of Twitter and the like, Blogs should be a more considered place for one's writing and bloggers should be putting more effort in to their work. A recent post I read went something along the lines of "....suppliers were trying their up most to...". Up most, from the word upperm...