Hi, I'm having serious issues with a 2003 Standard SP2 server (x86). Server is an hp dl360 g5, dual Quad Core Xeons and 4GB RAM and is being used as fileserver and is running Progress Database. There are about 40-50 users connecting, max 20 concurrent XP Pro clients and max 25 concurrent Terminal Server sessions from a 2008R2 x64 server. Server is heavily used as a fileserver. At this moment it will stop responding to clients every 2 days and the eventlogs will show Srv 2020 errors (unable allocate paged pool because it was empty). At this time i usually cannot connect through RDP and am forced to restart. If i do get onto it, i cannot start any program or copy any file without getting unsufficient resources error or i/o errors. This happened to the server a couple of months before when they were storing Outlook PST files on it, and i resolved it then by moving those PST files to a NAS. At this point i cannot move anymore data to another server I've already tried MS KB articles that corrspond to this issue (PagedPoolSize etc.) but to no avail. Does anybody have an idea? This is getting seriously frustrating for both me and my client.. Thanks! Dennes

Hi, Two Options: - Start performance monitor and monitor the processes and the page pool and check which process is consuming it. After you find the process deal with it. - Upgrade to x64 and add Ram to increase the paged pool. This will delay the problem but depending on how oft you receive it if the delay is he and you can afford regular reboots for maintenance for example to free the paged pool , you will be ok. Best Regards, Spas Kaloferov [ MCITP: SA6 | EA6 | VA7 | EDA7 |DBA10 | DBD10 | BID10 | EMA14 | SPA14 ] NetShell Services & Solutions | “Design the future with simplicity and elegance” Visit me at: www.spaskaloferov.com | www: www.netshell-solutions.com This posting is provided "AS IS" with no warranties or guarantees, and confers no rights.

Hi, Thanks for posting here. In addition, as a busy file server, it actually has a large demanding on the Paged Pool. Usually we recommend to do some tuning to enlarge the Maximum Paged Pool. Please follow the following steps to modify the key: 1. Run regedit 2. Locate to: HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\Memory Management 3. Add the following key: Value name: PoolUsageMaximum Data type: REG_DWORD Radix: Decimal Value data: 40 4. Exit regedit 5. Restart server to take into effect For more information on Paged Pool Tuning best practice, please refer to this KB: 312362 Server is unable to allocate memory from the system paged pool http://support.microsoft.com/default.aspx?scid=kb;EN-US;312362 Regards, Tiger Li TechNet Subscriber Support in forum If you have any feedback on our support, please contact tnmff@microsoft.com. Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.

Thanks for the comments. In addition, we were getting Srv errors 2021 (Description: Server was unable to create a work item %2 times in the last %3 seconds) in the system log yesterday afternoon, and lots of complaints from users experiencing very slow server/share access times. After opening a support call with MS, i was instructed to add these registry keys: - In HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager\Memory Management: PoolUsageMaximum (REG_DWORD, value: 60 decimal) PagedPoolSize (REG_DWORD, value: FFFFFFFF hexadecimal) - In HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Lanmanserver\Parameters: IRPstackSize (REG_DWORD, value: 4 decimal) MaxMpxCt (REG_DWORD, value: 2048 decimal) MaxWorkItems (REG_DWORD, value: 8192 decimal) MaxRawWorkItems (REG_DWORD, value: 512 decimal) MaxFreeConnections (REG_DWORD, value: 100 decimal) MinFreeConnections (REG_DWORD, value: 32 decimal) SizReqBuf (REG_DWORD, value: 16644 decimal) - In HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\Session Manager\Configuration Manager (create this subkey if it does not exist): RegistryLazyFlushInterval (REG_DWORD, value: 60 decimal) - In HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters: EnableTCPChimney (REG_DWORD, value: 0) EnableRSS (REG_DWORD, value: 0) EnableTCPA (REG_DWORD, value: 0) DisableTaskOffload (REG_DWORD, value: 1) - If not present already, install the latest version of UPHClean I also installed the latest HP PSP (driver pack) for the server, installed the latest Windows Updates and rebooted. As of this morning, the eventlogs seem to stay clean, with no more 2021 errors and i have a total used kernel memory of about 80MB, so everything seems to be running fine.. for now.. I'll just keep monitoring the server for the next few days and see if it stays this way.. Dennes

Well.. That was wishful thinking. Server is coming up with the dreaded event Srv 2021 again and the XP clients are all over me that their systems are extremely slow accessing the server shares and the redirected startmenu. Anybody have an idea?

Hi Dennes, Thanks for update. Just thought, have you recently modified any system settings before this issue occurred ? I’d suggest to test by booting in safe mode with networking to see if this issue will persist. This will help us to make sure this issue will not be affected by any third party software. How to troubleshoot Event ID 2021 and Event ID 2022 http://support.microsoft.com/default.aspx?scid=kb;EN-US;317249 Have you get any further suggestion for Microsoft support service ? Thanks. Tiger Li Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.

Hi Dannes, If there is any update on this issue, please feel free to let us know. Regards, Tiger Li TechNet Subscriber Support in forum If you have any feedback on our support, please contact tnmff@microsoft.com.Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.

Hi, Did you ever manage to find a fix for this? I'm experiencing almost exactly the same issue with a very similar setup.. Setup: HP ML350G5, Server 2003 Standard, 4GB Ram, 2 Xeon quad core cpus, mirrored OS, Raid 5 for everything else, over 50% available drive space across all drives: DC, DNS, DHCP, Exchange, file server. Serves approx 15 users through gig switches. Approximately every 2 weeks the server will become unresponsive, the only solution to restart. After which the server boots fine and works as it should. It can happen during the day/night/weekend/working day - random. This is not a new setup, it has been functioning fine for some 2/3 years, only in the last 2 months has the issue occurred. The first noticeable error is SRV 2020 Error (The server was unable to allocate from the system page pool because the page pool was empty). This will then flood the event log every few seconds and in the meantime many other errors are thrown up and services start to fail. Eventually the system will become unusable and clients will be disconnected with the error not enough server storage to process this command when trying to access shares. Once users are kicked off Its not possible to run any diagnostics of any kind on the server. You cant access task manager, services, event log, cmd prompt, nas drive, ping, get on line, browse the contents of the c drive etc etc All the time however network traffic (which you can bizarrely view) and disk activity look active. After a restart things go back to normal. You can then get access to the event log, however it becomes corrupt if you scroll back to the offending time when it starts to really go wrong. Main Events we see :: 2020 SRV :: 1026 MSExchangeSA :: 5005 MSExchangeSA :: 5007 MSExchangeSA :: 264 Foundation Agents :: 264 NIC Agents :: 4353 Foundation Agents :: 515 Storage Agents :: 1030 Userenv :: 1009 W3SVC :: 1002 W3SVC :: 7 KDC *A lot of these relate to a plain lack of resources though Things I have tried so far :: Memory test OK :: Checked Raid arrays / replaced drives OK :: Updated all drivers / patches :: Removed all non-critical software :: Stopped all non-critical services :: Followed MS KB for each event error (the list is too long but can give specifics if requested) :: Changed backup software :: Changed Antivirus software :: Virus Checked OK :: Checked for any scheduled tasks OK :: Increased Paging file :: Moved Paging file to another drive What's also strange is Ive been trouble shooting slow network throughout the week with a group of users. After a restart of the server to fix the unresponsive issue their network speed improves dramatically (from 160kb/s to over 200mb/s). Its almost as if the server degrades quietly over a period of time until it hits a limit where it just cant do any more and gives up. I'm guessing it is some kind of memory leak or faulty/buggy program but it's extremely difficult to pin down. Ive been trouble shooting this issue for 2 months now and have found no solution even close. I would appreciate any assistance or ideas anybody has. Reading through this post it seems very similar to the issues we have. I'd appreciate any assistance with this one. Regards Adam

Anybody have any ideas on this one? ::bump