First Day Back.
by, 5th September 2012 at 05:59 PM (1986 Views)
Wowsers. Sitting here now glad today is over in a sense.
It's not been a bad day - users thoughts seem largely positive or non-existent. NM made it very clear at a training day session that we've got a whole new network built from the ground up and there *will* be glitches, there *will* be niggles and there certainly will be missing bits of software. As and when you find those bits of software that used to be buried away deep in the depths of the old system, give us a couple of weeks to get through them!
It started off fairly badly yesterday morning when the print/papercut server fell over and croaked. It wasn't exactly a beast, it was an AMD X2 system running Server 2008 on Raid1 hard drives. Basically stuff we knocked up from spares a while back. We only found that out when we kept getting further printers not mapping correctly/at all and the server restarted itself a few times. So a new proper server was knocked up sharpish and finally got up and running.
So roll on today. The bulk of the problems were either printing or "Sims isn't on my machine/not working". A few copied connect.ini and SIMS installs later the majority were sorted throughout the day. Printing though kept cropping up randomly, and despite having copied printer names word for word the daft printer sharing system decided to choose names other than what I had actually typed in. Some head-scratching and adjustment later, that was solved.
It did become apparent though that perhaps our GPO per printer setup is a little unwieldy, but thankfully quick to apply. Mostly down to OU structure more than anything else - for instance it's easy to apply one printer to one room but it soon gets confusing when some of these rooms have a computer which needs to be in another OU , meaning you need to link the GPO to each particular OU. It's not difficult or complicated, but doubling things up. Perhaps when things calm down I'll move it to item level targetting instead.
Windows 7 is certainly a little harsher on networking than XP. Some of the cabling that's been in place for <long_period_of_time> had been fine before but caused lots of random problems with 7. Positively it gave us a chance to sort it out and maybe also solve some long standing performance issues it caused.
We also had some problems with permissions which we couldn't get our head round for a good deal of yesterday. We use DFS namespaces and replication across the two main servers for both the user directories and public/shared/staff folders. The users side of things works perfectly (so far) but people couldn't seem to get the right permissions to the shared or staff folders. All 3 of us had gone through share and NTFS permissions with a fine toothcomb, checked DFS settings were set to inherit NTFS permissions and couldn't find anything. Thankfully late in the day yesterday my colleague stumbled across Share and Storage Management console, and setting the permissions there sorted it out instantly. Still none of us know why that would be any different to doing share/ntfs permissions any other way, but as it works all we can do is leave it like that! We'll put it down to a foible of 2008 until we do know
Ah well, Roll on to the end of a very busy but ultimately positive day. Around about the time when one of us had set schedules to shutdown on Impero previously and entirely forgotten about it.
Thankfully he'd also been remoted into a couple of machines in time to notice the "You're being forced to shut down. You've got a minute to save your work."
Cue phone calls. Lots of phone calls. Bugger!
My colleague gets through amongst the panicked staff to be greeted with me dropping a cup of tea over myself and then the phone, followed by some brief swearing and laughing, of course she's wondering the same, noting that one of the servers had followed suit.
All 4 main servers had it too, and of course had no-one to press Cancel.
So, got those restarted and thankfully seems to have not caused a massive problem. One person had been working on a huge powerpoint all day and was very worried about losing it (although autosave should work, I'll check that tomorrow I think!) but once servers returned full access was restored and everything worked as normal. That was a huge relief and our first unofficial test of a full server failure which was great in a way - on CC3 when that sort of thing happened, clients had to be restarted fully before they'd ever behave again!
So a panicky end to a long busy day. I recall saying that if we got back to a manageable helpdesk in January then I'd consider this whole migration a success. I'm extremely optimistic and confident in that, although it is early days.One thing is for sure though - all 3 of us intend to have all or the majority of October half term off. It's well deserved, I think!
Find some time to relax.
Documentation, both for users and for ourselves/reference
Do some reading on the finer arts of SCCM
Keep Calm and Carry On!
Total Trackbacks 0