Sign in
 
HomeWindows Server 2012Windows Server 2008 R2Windows Server 2003LibraryForums
Windows Server TechCenter >
> 
> 
2008R2 - GPClient is taking a long time to handle the logon request (Up to 1 hour)

Unanswered 2008R2 - GPClient is taking a long time to handle the logon request (Up to 1 hour)

  • Wednesday, November 23, 2011 8:25 PM
     
     
    Sign In to Vote
    0

    Hi

    I had this issue a month ago and after spending a day trying to fix it I gave up and rebuilt the server. I had my RDS server built and functioning for 6 months until one day it started taking up to an hour for people to logon on to it, looking at the even logs the only event being logged was the GPClient is taking a long time to handle the logon request and this happend even when trying to logon locally. Rebooting the server didn't solve it, removing the network cables brought the server up so I could logon locally, however plugging the network cables back in instantly slowed the server down and went back to the same message in the event log of the GPClient is taking a long time to handle the request.

    Come forward a month and everything was working fine once again until yesterday where a user came to say that they couldn't logon. The message they were getting was "The task you are trying to do can't be completed because Remote Desktop Services is currently busy.  Please try again in a few minutes.  Other users should still be able to log on." Which was true, because I could logon to the server no problem. I decided to restart the server and the server hung (matching http://support.microsoft.com/kb/2601888). A Hard reboot later I am having the same issue as I was having last month with the event log logging the GPClient is taking a long time to handle the logon request. This time, I removed Sophos, disbanded the network team, uninstalled all the network cards from device manager, reinstalled the network cards, connected a single port on the internal network card to the switch, reallocated the static IP, reinstalled Sophos and the server started functioning normally again. It worked fine from last night until 12PM today at which point the logon time jumped from near instant to 20 minutend gradually getting longer, again the GPClient is taking a long time to handle the request. This time I rebooted my core switch and everything started to work again, apart from a few Group Policies being listed in the event log as causing problems and would not be reapplied until the server had been rebooted.

    Rebooting the server successfully applies the policy.

    The only common link between each time it has happend is it is the GPClient service causing the problem by taking a long time to handle the request. Occaisionally, my proxy settings aren't being applied by the GPO, but normally login off and back on again fixes this so it suggests something is going on with Group Policy, but how can I begin to work that out when it keeps working fine for ages?

    Server is fully upto date and has SP1 installed, as it stands at the moment I'm ready to smash it up!

    Any help, diagnostics, advice appreciated.

     

All Replies

  • Thursday, November 24, 2011 9:03 AM
    Moderator
     
     
    Sign In to Vote
    0

    Hi,

     

    I would like to confirm whether there are some logon scripts applied to this machine and the policy “Run logon scripts synchronously” is enabled. If so, please temporarily disable the Group Policy or remove the logon scripts to check the result.

     

    You may also refer to the blog below.

    The case of the mysterious 10 minute logon delay

    http://blogs.technet.com/b/instan/archive/2010/07/06/the-case-of-the-mysterious-10-minute-logon-delay.aspx

     

     

    Regards,

    Dollar Wang

    Forum Support

    Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.

     

    Technology changes life……
     
  • Thursday, November 24, 2011 3:50 PM
     
     
    Sign In to Vote
    0

    Hi

    There was one logon script configured for a group of users in 1 OU which I have recently disabled, the only logon script that is now configured is the one that is set through the profile tab on the users account.

    Run logon scripts synchronously has not been configured so is at it's default value of disabled.

    I'll have a look through the blog and see if I can spot anything else.

    Thanks

     
  • Thursday, November 24, 2011 7:54 PM
     
     
    Sign In to Vote
    0

    Hello,

    i have a similar problem. When i reboot a rdsh (Windows 2008R2) it takes along time  (one hour) for the first logon, when i use the console.

    There is no rdp connection possible.

    You can reach the admin share, also connect to eventlog and services remote.

    I think, this is also a problem with group policy (prefernces) we use.

    i have found the following hotfix http://support.microsoft.com/default.aspx?scid=kb;EN-US;2561285

    Do you have installed this hotfix? Do you have installed sp1?

    I have "cloned" the rdsh with sysprep, all servers running on hyper-v.

    Regards

     
  • Thursday, November 24, 2011 10:08 PM
     
     
    Sign In to Vote
    0

    Hello,

    i have a similar problem. When i reboot a rdsh (Windows 2008R2) it takes along time  (one hour) for the first logon, when i use the console.

    There is no rdp connection possible.

    You can reach the admin share, also connect to eventlog and services remote.


    This is exactly the issue we have too. All remote management works fine.

    I hadn't spotted that hotfix, I'll download it and give it a shot. We do have Service Pack 1 installed.

    @Dollar Wang: I've read that blog post and spotted it mentions that the Notification time out has increased to 10 minutes, the event log definately has logged GPClient as taking 3599 seconds to handle the notification.

    Touchwood, today hasn't been a problem.

     
  • Friday, November 25, 2011 7:07 AM
     
     
    Sign In to Vote
    0

    Hello,

    i have also installed sp1. The problems was occuring also before sp1 installation.

    Sometimes, after reboot, no problems.
    Server comes up, applying the computer settings, string+alt+del screen. Everybody can login. 

    I am also thinking about a call by microsoft. But i can not really reproduce the problem every time. So it will be hard to troubleshoot.
    Anybody knows a way to collect startup indformation?
    Maybe there is a hint for the problem. Or i can use this information when i open a call.
    The event log has no errors which will help.
    For me it looks like "the system is waiting for something".

    I have also not installed the hotfix. Will install next week.

    Regards

     

     
  • Tuesday, December 06, 2011 4:10 PM
     
     
    Sign In to Vote
    0

    Hello,

     

    i have installed the update. Did not help. But we have enabled a policy "wait for network". Maybe there is a problem.

    @ijefferson: Do you have the policy ""Always wait for the network at computer startup and logon" enabled?

    Could this be a problem? Maybe the server is waiting, waiting until time out.

    Regards

     
  • Tuesday, December 06, 2011 5:20 PM
     
     
    Sign In to Vote
    0

    Hi

    I haven't had time to install the hotfix on my server yet, my maintenance cycle isn't until tomorrow.

    I don't think I have that policy set for the Terminal server, I'll confirm when I'm back in tomorrow.

    Strangely, the server has been working without issue since I posted this thread.

     
  • Thursday, December 08, 2011 3:19 AM
    Moderator
     
     
    Sign In to Vote
    0
    OK, looking forward to your feedback.
    Technology changes life……
     
  • Thursday, December 08, 2011 9:33 AM
     
     
    Sign In to Vote
    0

    As the hotfix didn't make any difference to you, I haven't installed it and the issue still hasn't reoccured yet (although I fear it's only a matter of time before it does).

    I haven't got the "Always wait for the network at computer startup and logon" policy enabled either so that half rules out that possibility out as well I suppose.

    Still very confused over it.

     
  • Thursday, December 08, 2011 2:56 PM
     
     
    Sign In to Vote
    0

    Hello,

    We have a very similar issue to the one described above (GPClient), we are about to deploy the hotfix to our problem Terminal Servers 2008R2, however do i have to apply this hotfix to the Domain Controllers as well or just the client computers?

    Cheers

    Mark

     
  • Wednesday, January 04, 2012 8:56 PM
     
     
    Sign In to Vote
    0
    @Marc - How are you progressing with this issue? Touch wood, so far it hasn't reoccured again for us (which is what makes it so annoying to try and fix), but we were closed for 2 weeks so the server hasn't had a huge ammount of use.
     
  • Wednesday, January 04, 2012 9:05 PM
     
     
    Sign In to Vote
    0

    hello,

     

    hotfix does not help...

    I have restarted some server and the failures occurs again. Not every time. So oit is very hard to find.

    Last time, i was able to logon with local administrator account. Then i have seen a circle over the networking symbol next the clock in the task bar.

    After an hour the circle was gone and the server ok. Before ping ok, rdp port not available. Also our monitoring was not able to reach the server to check services and so on....

    Really strange... :(

    Every server is running on hyper-v. Only terminalserver farm has this problem. But, we have 3 farms. Only one farm has the problem.

    Other servers has no problem.

    Regards

     
  • Friday, January 06, 2012 1:13 PM
     
     
    Sign In to Vote
    0

    Hello,

     

    today i have restarted one server 3 times. Every time same problem:

    i can only login with local administrator. i can ping devices in the network. i can also access network shares. i can ping the server from other devices. but rdp port is not enabled. also access to wmi is not possible.

    i have a blue circle on the network icon in the tray. After a specific time i get a red x. But i can query dns, surfing.

    i can open the "network and sharing centre". I can also get the status of the nic. But when i try top open the properties nothing happens.

    Hyper-v guest, so there is no special network card or drivers. I have deinstalled the nic and restarted the server. NIC was detectet and automatically installed.  But this did not help...

    Anybody an idea?

    Regards

     
  • Monday, January 09, 2012 9:38 AM
     
     
    Sign In to Vote
    0

    Hello,

    another update: Today same probem. I have logged in with local admin credentials, after that i have seen in taskmanager svchost is using 50 percent cpu and around 600 Megabytes memory.

    With sysinternals tool i have seen, it is the svchost which is running group policy, windows auto update, profile server and so on.

    With this information i have found http://support.citrix.com/article/CTX128749  (thanks to citrix)

    I have found a lot of files in the group policy history "c:\programdata\microsoft\group policy\history" (hidden) folder. After deleting i have restarted the server 8 times. Every time performance great. Login possible without long waiting time...

    Maybe the group policy service is scanning/reading the folder or anything else.

    @ijefferson: Maybe it will help also in your environment...

    Next weeks will show if it really helps.

    Regards

     

     
  • Monday, January 09, 2012 10:41 AM
     
     
    Sign In to Vote
    0

    Funny, the same issue has literally just occured for me now but yet 30 minutes before I was using the server without issue.

    I'm just trying the history folder now and will see if that sorts the issue (my trick of last time uninstalling the network cards hasn't fixed it).

    I'll report back....

     
  • Monday, January 09, 2012 11:22 AM
     
     
    Sign In to Vote
    0

    Right, I checked the history folder and I had 4 folders within there and within each one of those hundreds as you say (all guids, so for each user as per the citrix article).

    Deleting that and restarting made no difference to me, so I went back to process monitor and noticed that there was a service hogging all the memory, and it was the Windows Management Instrumentation service. When I looked, it spent a long long time reading the repository folder and then other services started to kick in after a bit, but still very slowly and the RDS service still wouldn't allow others to log on. When I eventually got the WMI service stopped (The IP Helper service hung first, which after it went into a stopping state, I could End the process and the same with the WMI service), I could rename the repoistory folder. I've rebooted and everything is working properly again now.

    The current size of the Repository folder is now 16MB, the one I renamed was over 600MB!

    @Marc - Out of interest, how big is your repository folder? (c:\windows\system32\wbem\repository)

     

     
  • Monday, January 09, 2012 11:30 AM
     
     
    Sign In to Vote
    0

    Hello,

     

    i have also tried to stop ip helper. No chance. After timeout the service "was stopping".

    My wmi rep is about 580 MB. Another Win2008R2 server, not RDSH, has about 60 MB. Installation also 1,5 years ago.

    Could also a problem for us. Biggest files are objects.data (430 MB) and index.btr (about 100 MB).

    Do you use monitoring software with wmi queries?

     

    Regards

     
  • Monday, January 09, 2012 11:50 AM
     
     
    Sign In to Vote
    1

    @ijefferson:

    maybe WMI hotfixe will help, i will install them next planned downtime...

    http://social.technet.microsoft.com/wiki/contents/articles/1398.aspx

    There is http://support.microsoft.com/kb/2505348/en-us or http://support.microsoft.com/kb/982293/en-us and more

    http://blogs.technet.com/b/askperf/archive/2011/08/05/suggested-hotfixes-for-wmi-related-issue-on-windows-platforms.aspx

    Regards

     

     
  • Monday, January 09, 2012 12:05 PM
     
     
    Sign In to Vote
    0

    I don't monitor anything with WMI myself, I think there might be one installation agent which uses WMI but I'll need to double check.

    I've got some maintenance planned later on this week so was thinking I might try 982293 and 2617858 (982293 especially because it mentions both the User Profile Service and Group Policy Client service).

    The 2 files you mentioned are the same as ours, index.btr was 201MB and Objects.data was 446MB.

    • Edited by ijefferson Monday, January 09, 2012 12:42 PM
    •  
     
  • Wednesday, January 11, 2012 6:41 AM
     
     
    Sign In to Vote
    0

    Hello,

    will install this hotfixes also later this week. Hope, it helps.

    Maybe you can update this thread after installing?

    Thanks

     

    Regards

     
  • Wednesday, January 11, 2012 5:06 PM
     
     
    Sign In to Vote
    0

    Hi Marc

    I've installed 2617858 tonight and it seemed to go ok, no problems following the restart. I tried to install 982293 but just not a message saying that it wasn't applicable to the machine so a no go with that one unfortunately.

    Will let you know how things progress.

    Ian

     
  • Friday, January 27, 2012 6:49 AM
     
     
    Sign In to Vote
    0

    Hello,

     

    i have installed all patches and cleared the gpp history folder on different servers. At the moment after a lot of restarts the logon delay did not occur again.

    Thanks

    Regards

     
  • Tuesday, January 31, 2012 12:01 PM
     
     
    Sign In to Vote
    0

    Hi Marc

    How are things going on your end?

    Since applying the hotfix and making a few more changes the issues haven't reappeared (I say very tentatively). However because I made more changes I'm not sure whether it's just biding it's time, the hotfix fixed it or the other changes I made have sorted it.

    If it's reoccured for you I'm guessing the hotfix won't have worked, in which case maybe we could try the other route I've gone down and try an compare notes.

    Ian

     
  • Tuesday, January 31, 2012 3:06 PM
     
     
    Sign In to Vote
    0

    Hello,

    at the moment, the start behavior is normal. I have installed the hotfixes and deleted the history gpp folder.

    svchost.exe is working normally after restarting. I have tested it with 5 different servers (all servers was affected by the long delay after restart) and at the moment everything is "normal".

    I will add the "delete cache folder on shutdown" to my servers. Hope the problems are gone.

    Regards