Server 2012 File Server - suddenly stops serving requests (but otherwise looks fine)

[sted]

I personally think it's update related. My 2012 file server has only the File and Storage Services role installed and hasn't had any updates applied since August (I know, I know!) and has never suffered from this problem.

 

ditto. there were some patches for 08r2vthat caused me headaches earlier in the year so for the moment ive stopped updating servers and 2012 auto reboot is silly you donparches forget to reboot then it just "randomly" reboots itself to complete bad enough on workstations dumb on a server

[Ste_Harve]

I think if it were me, I'd be inclined to put a 2008 R2 box back on the network and move shares back. It clearly doesn't suffer from this issue and it's weird Microsoft are being rather quiet about it. As I say, I hope all of you with this problem have reported it!

 

Thats my plan for now, but need to find time to migrate shares as don't use DFS!!

[JamesDonkin]

Having simillar bother here on three virtualised (hyper-v) 2012 servers. Been having people going online / offline randomly. Noticed something in SMB client logs on the servers though.

 

Event ID: 30621 SMBClientSession to server \servername was lost Status 0xC0000203

 

Event ID: 30623 SMBClient

Connection to share \servername\Data was lost. Status 0xC0000203

 

And then they instantly reconnect. This happens several times a day. Think its linked?

Edited December 10, 2013 by JamesDonkin

[CHR1S]

I personally think it's update related. My 2012 file server has only the File and Storage Services role installed and hasn't had any updates applied since August (I know, I know!) and has never suffered from this problem.

 

My unpatched 2012 servers did this during the setup stage, keep clutching haha

[JamesDonkin]

My unpatched 2012 servers did this during the setup stage, keep clutching haha

 

Did you apply any patches other than those on windows update?

[CHR1S]

Nope, I didnt patch at all before it happened. I was in the process of setting up a WSUS server so just waited until that was up and running.

[JamesDonkin]

I was excited about some kind of magic bullet there but they're all up to date by windows update, cheers anyway though. Our main storage is a SAN on iSCSI which I'm going to give its own NIC x-over cable and subnet tonight. See if that has anything to do with it although I doubt it, I'm clutching at straws.

[lmgtfy]

Well in some respects I'm glad I'm not the only one with this issue it's now going to be the first brave sole who waits for the issue to present and then gets in contact with Microsoft when's it's not working who is going to win this battle. Main problem is who can afford to do this when a restart solves it and ringing Microsoft and getting any useful helps takes an age in my experience anyway.

[psydii]

Well in some respects I'm glad I'm not the only one with this issue it's now going to be the first brave sole who waits for the issue to present and then gets in contact with Microsoft when's it's not working who is going to win this battle. Main problem is who can afford to do this when a restart solves it and ringing Microsoft and getting any useful helps takes an age in my experience anyway.

 

Assuming you are fully up to date with Windows Server update rollups, and recommended windows 7 hotfixes,

log the call before it next manifests, state that it has happened before.

 

MS can show you how to get your server to produce a crash dump which they can use for debugging, it only takes marginally longer than a straight up reboot.

 

Also the new message analyser tool may be able to do circular logging, and therefor perhaps have tracing and packet captures available of the machine/network state in the moments before srv2.sys turned up its toes... So opening a case with MS probably isn't a waste of time.

Edited December 10, 2013 by psydii

[jamesfed]

Just revisiting this post - had what I would call a 'slow down' today on the server that I migrated all of my data from (which was having this problem origoinaly).

 

I noticed that requests to the server for files were taking much longer than usual, a quick reboot soon sorted it out.

 

So simlar but not the same....

 

Anyone had any new thoughts?

[TheCrust]

Just seen this thread.

 

Put a new (well, repurposed hardware) server in over the Christmas break to use as a staff file server:

 

• 2012 R2 Server Standard Edition. Running DNS server and ADDS roles (as the previous 2008 R2 box was)
  
• HP Proliant DL380 G5: dual quad-core Xeon, 24Gb RAM, teamed NIC, current firmware.
  
• Files stored on external HP MSA2012sa array, directly attached to server rather than via iSCSI.
  
• FSRM role installed. Quota monitoring (not enforcement) on staff home drives, but no other folders.
  

 

About once a day, access from my Windows 8.1 Enterprise client just hangs: can't open email attachments, direct access to files stored on this server stops either through Explorer or individual applications, can't open Explorer as takes an age to refresh view of mapped drives. Interestingly I noticed existing explorer windows still display files OK and can move between directories on the server if they were open at time behaviour occurs.

 

The server happily returns a ping with normal response times, and can be RDPd into when it has one of these moments.

 

A reboot of the workstation seems to sort access out for me and as none of the other Win 8.1 or 7 clients seem to have this problem - at least nobody else has yet mentioned this sort of behaviour to me - I had been assuming it was a problem on my individual workstation (the event logs on indicate possible drive corruption so I was going to blitz it and reinstall in a free moment).

 

At least, I was until I read this thread anyway!

 

As others have had what sounds like a similar fault, with Windows 2012 and 2012 R2, but there is little information as to what might be causing it, I may just log a call with our software support providers (ESC UK) on Monday as they will escalate up to Microsoft if needed.

Edited January 11, 2014 by TheCrust

[TheCrust]

Has anyone tried clearing down their profiles and starting again?

 

It's just a thought, but further to the above I rebuilt my Windows 8.1 client yesterday with a new hard drive. That seemed to speed the system up nicely, so I guess the old install wasn't particularly great (it had been an 8.0 install, upgraded to 8.1 when it went RTM and its not beyond the bounds of possibility that the new drive I'd fitted to install 8 was also an early failure), but the problem was still there - only worse. It would now take an age to show Computer if Explorer was opened for about the first five minutes after logging in, in addition to the regular cease-and-desist behaviour it would show randomly when accessing file shares on this one particular 2012 server.

 

Cleared down my profile first thing this morning and let Windows regenerate it as part of the next step of troubleshooting, and the whole lot seems so much happier. It hasn't had one of its moments yet today - though I've not been at my desk much today to be completely sure.

 

Has anyone else tried this step if the problem is consistent, but not necessarily reproducable on demand, with a user?

[bon_clay]

Happened to our 2012 server twice one last friday and the other time a couple of weeks before the Christmas break.

Had 2012 running for a couple of months as a file server as a non DC and the problem did not occur. A couple of days after we promoted it to a DC the problem occurred.

Does everyone else have there 2012 as a DC when this problem occurs?

[Michael]

Happened to our 2012 server twice one last friday and the other time a couple of weeks before the Christmas break.

Had 2012 running for a couple of months as a file server as a non DC and the problem did not occur. A couple of days after we promoted it to a DC the problem occurred.

Does everyone else have there 2012 as a DC when this problem occurs?

 

That's interesting - I never thought about the possible roles a server is hosting could cause the problem. So presumably you have it setup as a DC, running DNS, DHCP and File Services? Any other roles? Would be good to know what other roles the rest of you are running with this problem.

[bon_clay]

That's interesting - I never thought about the possible roles a server is hosting could cause the problem. So presumably you have it setup as a DC, running DNS, DHCP and File Services? Any other roles? Would be good to know what other roles the rest of you are running with this problem.

Current roles on the server are Active Directory Domain Services, DNS and File and storage services.

[3s-gtech]

Our 2012 file servers aren't DCs, and have been fine aside from one isolated incident. Our 2012 R2 DCs aren't file servers, and have been fine. So could be correlation there...

[Mr.Ben]

I saw this a few times in the first week of running 2012 (Hyper-V Clustered File Servers Here), but updates seemed to make it go away.

 

Yesterday we had the first instance of it since last September, then 20 minutes later the second instance. As with everyone else nothing immediately looked wrong.

 

Straw Clutching - Looking at our login logs, a staff member logged themselves into a whole room with Impero for a training session, and the first failure came 5 minutes after that?

 

No logs or useful information though!

[spc-rocket]

There is a list of hotfixes that are related to files services on Microsoft website.

 

List of currently available hotfixes for the File Services technologies in Windows Server 2012 and in Windows Server 2012 R2

 

Might be worth doing some tests with this. We don't have any 2012 r2 servers yet.

 

Ash.

[Gibson335]

That's interesting - I never thought about the possible roles a server is hosting could cause the problem. So presumably you have it setup as a DC, running DNS, DHCP and File Services? Any other roles? Would be good to know what other roles the rest of you are running with this problem.

 

Ours was just a data server, just student data...rolled back to 2008R2 and it's been fine.

[CHR1S]

Not had an issue since the first and only time. "Touch Wood"

[Homer]

We've had this two or three times since moving our FS services to 2012 - server slowly stops responding to file requests, and then stops entirely - at which point the phones light up! Responds to pings fine, remote desktop is alive and everything seems fine, but file services are down!

 

Reboot fixes it. Once I put this down to Veeam, as another technician hadn't ended a restore job for 2 days, and then somebody tried to do another one. At this point the server shart itself.

 

Any of you running Veeam?

[jamesfed]

Yep running Veeam here but never seen a backup do anything like that.

[Gibson335]

Nope - no Veam here.

[Homer]

Kind of rules Veeam out then... it's an odd issue, touch wood it's not shown up again recently so might have been fixed in a hotfix?

[TheCrust]

I'm still having issues with mine - clearing down my profile did not solve it, but I seem to be the only one that's affected by this issue at the moment. On average about once a day though the timing always seems different. Everyone else seems to be fine, or hasn't noticed if there's a problem.

 

I opened a call with our support company (who should refer it up to MS if they can't fix it).