[MonsterMaxx]

doesn't it always happen like this...I'm getting ready for a 2-3 week overseas business trip and on my very last day when I have a zillion things to do the mail server takes a poop.


It's a win2003 server w/ exchange and Intel gigabit nics.

This morning it was running, but had knocked a few shares offline so I rebooted it.
Nothing in networking is working now. All it lists is that the cables are unplugged. Yet NOTHING changed physically except a M$ update and a reboot. The cables are NOT unplugged and I've even switched a cable from a machine that's working. This is absolutely NOT in my network, it's in this bloody machine.

Prior to this I'd installed a few M$ updates, but hadn't rebooted.

This had happened to the sister machine (identical hardware) several months back and all I did to fix it was uninstall the M$ update.

This time no luck.

I'm totally screwed. Any bright ideas how to fix this piece of _ _ _ _?

This post has been edited by MonsterMaxx: 13 January 2009 - 11:14 AM

[cluberti]

Can you roll back the updates you installed?

[MonsterMaxx]

did that, no dice

I've removed all updates since november, reinstalled the nic drivers,

now I removed the nic team and am rebooting

One oddity is that it would not let me remove the nic from device manager and rescan for it.

Another oddity is that it says that the hardware has significantly changed and now it wants to reactivate. NO CHANGES in hardware

This post has been edited by MonsterMaxx: 13 January 2009 - 11:26 AM

[cluberti]

If rolling the updates back doesn't fix it (also, what KBs are we talking about here?), then you'll have to go back to your changelog to see what else happened before the reboot. I get a lot of questions like that (I installed MS Update <x> and rebooted, and now <y> is broken!). However, maybe 1 out of 100 does the problem actually go away when the update <x> is removed - thus it was something else that took effect after the reboot, and having a changelog is a good way to track these down. If you don't have it, I might suggest booting in safe mode (w/out networking) and remove the devices that way, then reboot and reinstall drivers.

Also, if you're getting complaints about major hardware changes, can you think of anything that would've installed / updated filter drivers (like antivirus, firewall, security CSPs, etc) that might have been installed or updated during the last uptime? That's the usual cause I see of this sort of behavior (device manager, networking problems, and need to reactivate / hardware issues).

[MonsterMaxx]

This machine just sits and runs. It's a mail server. Nothing new has been installed in ages.

It's fiercely p***ed off now. I did a hard shutdown and the power supply refused to come back up. Luckily the sister machine got a new case a year or so ago so I have the PS from it's case. Put it in, machine fired up and is now sitting there (for the last ten-fifteen mins) "preparing network connections" I have no doubt the problem is still there.

If this reboot doesn't help I'll try disabling the onboard dual port nic and use a PCI one.

I was able to delay the my trip by one day, hopefully one extra day to deal with this BS will be enough.

I don't remember for sure, but it seems the last time I had a fully screwed domain controller (3yrs ago) Safe mode wasn't an option. yes/no?

[MonsterMaxx]

I got it running, even put the M$ updates back.

No idea what the original problem was, though I did see a lot of complaining about M$ updates and the nic drivers.

I'm guessing the nic diver was a little old and this caused some incompatibility with M$s updates. I had to rip all that out, delete the teaming, start with the latest nic driver and put it all back. 15 reboots and a day later it's rebuilding a RAID. Another 10 or so hours of that and I can put it back in the rack.

What a pain in the backside.

When I get back from this trip I'm going to find a hosted Exchange or Open-Xchange service. I'm tired of screwing with mail servers. This little f-up cost me more than a year's hosting would.