Network gone doolally

[Simcfc73]

Hi guys.

 

My networks gone a bit stupid today. Last night I rebooted my exchange server and it didn't boot up right. I think it had something to do with a disk cleanup that was run which cleaned some files it didn't like (or I deleted some tmp files that shouldn't have). Anyway it wouldn't boot unless I removed the etherent cable (this is a rack server so it was a pain), and I managed to get it booted up and all the services working.

I went home happy(ish)

 

Today alot of the pupils can't log on. It hangs at loading personal settings and stays there. Now I've removed the GP and it loads straight up but I get a network error when it tries and writes the profile back to the server. So I thought it was a netowrk error. For some reason I thought it might be the exchange server, no it wasn't cos I removed that.

 

This seems to happen in rooms more than a few hops from the server which i would have hoped would have narrowed it down but it hasn't as I'm getting the error in different places. The weird thing is that 2 machines next to each other which have been cloned will load up at different speeds.

 

This is annoying the hell out of me cos I can't work out of its a hardware issue, or a software one.

 

I'm rebooting all the switches before I leave tonight hoping that it will be nice and smooth tommorow but if it isn't then can someone offer some advice.

 

I think I'll go home and see if I can get some more netowrk tools and see if there's anyting flying about that shouldn't.

 

Many Thanks

[Netman]

Hmmm...

Had a similar thing yesterday pm. Servers and clients hanging all over the network as if there was some massive broadcast storm going on... spent hours rebooting servers, switches, running AV and checking event logs with nothing untoward showing up - then it cleared and has been fine since with not a sniff of any related errors in logs anywhere....

makes ya wonder....

[ChrisH]

Ive had similar problems today as well. Im currently in the middle of migrating roles to new servers spreading the load by they are just running side by side at the moment where possible.

Basically one of the old servers is a DC/DNS and the file server. When this machine goes off it effects everything for some reason. Even down to freezing windows when its loaded!! I put this down to the redirected startmenus and default profile maybe which should be fine as its in the netlogon directory and theres a perfectly healthy (2 infact) DCs to serve the clients.

I changed DHCP so none of the clients are looking at that machine for DNS, I moved the Startmenus to a share on another machine and updated GP but the machines were still taking ages to log on!!!! It seems to me something is taking far too long to time out.

Anyway currently my file server is serving files in safe mode with networking :? while I try and figure out why it wont start properly.

Between that and wrestling a 42U rack into a room with a narrow coridoor I have had lots of fun today

 

Anyway I will step up my efforts and seize the roles from this server tommorow and probably reinstall server on it as a member server just serving files.

 

I just find it very strange in a so called multimaster server scenario everything goes tits up when that server isnt up. I mean I would expect a small timeout period and them to be able to log on with no start menu and not be connected to their home directory.

I know someone is going to scream its DNS but Im failing to see where when I have 2 other healthy DNS servers and that srver is no longer looked at for resolution.

[marco84]

We also had a similar problem today and yesterday, when a user logged in it was taking an age for the re-directed start menu to load. The problem seems to have corrected itself for now, i did have ethereal most of the day and didnt detect anything strange on the network.

Its all very strange. . . .

[Quackers]

Your not the only one, since yesterday my clients have freezed for a few seconds when loading the desktop after logging on, and seem to take longer.... is there a time bomb is MS's code lol ?

[Sylv3r]

Us too. Network basically grinded to a halt about 10:20 yesterday. Waited till breaktime before rebooting the servers and everything seemed fine after that.

[john]

WE have had some as well, I say its a MSFT or Sophos thing, probably Sophos as I blame everything on Sophos!!!

[Simcfc73]

Back up and running.

 

I rebooted everything, including the switches this morning at 7am and it looks okay now.

 

Very very weird.

 

Only thing that I noticed was unusual was in the event log when I rebooted the main server. A DCOM error to do with Sophos (again) which was running under the admin account which I fixed and the network card showed it turning on and off 4 times just before it loaded all the services. This has never happened before so I enabled the other network card int he server and switched that on too. No harm in running two. I don't fancy bridging them just yet though just in case its a faulty card.

 

Anyway, the funs over, back to resetting the kids passwords and finding work that they've 'lost'

[Gatt]

FRAK!! our network is doing the same thing AGAIN - roughly the same time too!

 

Anyone else getting it?

[webman]

Must be electromagnetic interference from the inter-dimensional time portals that were opened up by Mars' present position in the solar system

 

I actually have no idea why it's happening, but I can say that we aren't having any problems, but we don't run Sophos. Probably not much help. :oops:

[Gatt]

Well ours went Tits up - big style - we lost our PDC completely, so had to seize all the roles to the BDC

Rebooted all the servers, then had problems with the kids PC no longer seeing the domain and unable to log in

 

Eventually rememebered that the PDC (deceased) had the DHCP scopes and DNS entries!! so had to recreate all them

 

Still couldn't get PC's to log in, eventually found that we had to unjoin them all from the domain, and re add them back again!

 

On top of that staff took this opportunity to fire stupid faults at us (ie SIMs not working, can you setup an email account, I;m out of Ink!! :twisted: )

 

They'll have to wait!

[beeswax]

I don't think it's Sophos, unless Moorside has changed its AV software. Does everyone who has posted to this site use server 2003 with XP Pro clients? Is there a common thread here? A new patch or update perhaps. All this within the space of just over 24 hours.

Pass me the meerschaum Watson, I think this is a 3 pipe problem.

 

beeswax

[Gatt]

Yeah Win 2k3 SP1 & XP Pro Sp2

Now using Mcafee 8 for AV

 

Was wondering if there was something common here too, we were getting a LOT of problems with the PDC recently tho with Replication Errors, DNS errors so reckon there was more too it.

 

Just sods law that it started during an HMI inspection, but thankfully it didnt go tits up till they left!!

[john]

XP Pro SP2, all latest updates through WSUS, Server 2003 SP1 and again all updates. Its freakey, but it didn't happen today I don't think, although I was in a meeting between 9:30 and 10:30 which is when it occured for me yesterday

[Simcfc73]

2003 SP1 here with XP SP2 all around.

 

How annoying is it not knowing why.

 

I may go to Linux.

 

HAHAHAHA

[SteveT]

We have also had strange goings-on this week.......permissions changing on some files etc. Only thing I can think is it`s something to do with KS3 IT test..........it`s the only change carried out recently!!! Maybe an attempt by RM to boggle vanilla windows networks

[Ueline]

Is machine citys collective unconscious stirring Mr Anderson?

[E1uSiV3]

hrmm interesting, i havent has a single problem here and i have very similar setups, the only thing different being i am no longer at a school so am not running the KS3 pilot...

[beeswax]

We have also had strange goings-on this week.......permissions changing on some files etc. Only thing I can think is it`s something to do with KS3 IT test..........it`s the only change carried out recently!!! Maybe an attempt by RM to boggle vanilla windows networks

 

we haven't had any noticable problems, but we're only part way through running the KS3 pilots. but all server2003 and xp pro.

beeswax

[Netman]

No KS3 pilot... No Sophos (Panda instead) and the same thing happened again today.... lots of failed security events in the log but no indication of why... Nearly 2 hours with all clients and servers hanging... Aaarrggghh..

[SteveT]

We have also had strange goings-on this week.......permissions changing on some files etc. Only thing I can think is it`s something to do with KS3 IT test..........it`s the only change carried out recently!!! Maybe an attempt by RM to boggle vanilla windows networks

 

OK......still having problems, none of which were apparent before installing KS3 Test software. Can anyone think of any other reason for the following:

 

1. Permissions on pupils user areas being changed randomly.

2. Files/folders disappearing from various areas......including shared areas.

3. This seems to take place overnight or over the weekend so not a deliberate act by anyone on our network.

4. Nothing untoward in server logs.

5. I`m aware that the ICT test software carries out various "housekeeping" tasks overnight........DB Vacuum makes me worried!!!!

 

If no-one can think of any other explanation for this I think it`s time to remove this software from our system!!!!