none
Application Pool problems RRS feed

  • Question

  • Hello, I use SCOM to monitor my Exchange Server 2013 DAG (2 nodes) on Windows Server 2012. The issue we are experiencing in one of the nodes is that all app pools are being recycled from time to time. I get several warnings in SCOM with title "Application Pool worker process is unresponsive" and Application Pool worker process terminated unexpectedly. In event viewer I also get this warning, one per app pool:

    Log Name:      System
    Source:        Microsoft-Windows-WAS
    Date:          22/09/2014 14:57:42
    Event ID:      5013
    Task Category: None
    Level:         Warning
    Keywords:      Classic
    User:          N/A
    Computer:      ex02.xxx.local
    Description:
    A process serving application pool 'MSExchangeRpcProxyAppPool' exceeded time limits during shut down. The process id was '6556'.
    Event Xml:
    <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
      <System>
        <Provider Name="Microsoft-Windows-WAS" Guid="{524B5D04-133C-4A62-8362-64E8EDB9CE40}" EventSourceName="WAS" />
        <EventID Qualifiers="32768">5013</EventID>
        <Version>0</Version>
        <Level>3</Level>
        <Task>0</Task>
        <Opcode>0</Opcode>
        <Keywords>0x80000000000000</Keywords>
        <TimeCreated SystemTime="2014-09-22T17:57:42.000000000Z" />
        <EventRecordID>35377</EventRecordID>
        <Correlation />
        <Execution ProcessID="0" ThreadID="0" />
        <Channel>System</Channel>
        <Computer>ex02.xxx.local</Computer>
        <Security />
      </System>
      <EventData>
        <Data Name="AppPoolID">MSExchangeRpcProxyAppPool</Data>
        <Data Name="ProcessID">6556</Data>
        <Binary>
        </Binary>
      </EventData>
    </Event>

    I hace recently reinstalled the whole server and later installed CU6 but the problem remains.

    Thanks in advance

    Gonzalo.


    Tuesday, September 23, 2014 12:49 PM

Answers

All replies

  • Hello,

    Thank you for your post.

    This is a quick note to let you know that we are performing research on this issue.

    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com


    Simon Wu
    TechNet Community Support

    Wednesday, September 24, 2014 2:11 AM
  • Hello Gonzalo,

    We would receive the event ID 5013 if one of the following is true:

    1. The ShutdownTimeLimit is set to too low of a number
    2. A problem has occurred inside of the web application that is hindering it from shutting down in a timely manner.

    The default ShutdownTimeLimit in IIS is 90 seconds, so please checking this parameter value first or we could  increase the “Shutdown Time Limit” to 200, do iisreset and see if the warning still occurs. Here’s a nice article telling us how to modify this value: http://www.iis.net/configreference/system.applicationhost/applicationpools/add/processmodel

    Another article regarding to this event:

    http://support.microsoft.com/kb/2634635/en-us

    Thanks


    Regards, Eric Zou

    Wednesday, September 24, 2014 6:41 AM
  • Hello Eric, thank you for your answer. I checked the ShutdownTimeLimit value and it is 90 indeed. I also have te same value in the other DAG member and there aren´t problems in this one.

    I also receive this warnings in event viewer at the same time:

    ID 5011 Source WAS:

    A process serving application pool 'MSExchangeECPAppPool' suffered a fatal communication error with the Windows Process Activation Service. The process id was '21272'. The data field contains the error number.

    ID 5138 Source WAS:

    A worker process '6556' serving application pool 'MSExchangeRpcProxyAppPool' failed to stop a listener channel for protocol 'http' in the allotted time.  The data field contains the error number.

    Do you still recommend to extend the shutdown time limit?

    best regards

    Gonzalo

    Wednesday, September 24, 2014 12:42 PM
  • Hello Gonzalo,

    Yes, please extend the shutdown time limit and do IISRESET then monitor the issue.

    Meanwhile, could you please help to answer the following questions to narrow down the issue:

    1) Since the issue only happens for one server, does this affected server handle most requests from clients?

    2) I noticed the event 5103 and 5138 all mentioned the process ID "6556", and the event "5011" indicates the process "21272", maybe we could run get-process in Windows PowerShell or check it in Task Manager to find the corresponding instance.

    3) Does this a MBX&CAS all-in-one server?

    4) NLB deployed?

    5) Do any real impact for the end users? Such as Outlook disconnected.

    Moreover, please check the following IIS related article to see it fits our condition:

    http://support.microsoft.com/kb/831135

    Thanks.



    Regards, Eric Zou


    • Edited by Eric ZouZou Thursday, September 25, 2014 9:19 AM
    Thursday, September 25, 2014 9:19 AM
  • Eric, I configured the recommended value to 200 and then iisreset. I will monitor the SCOM console to see if the warnings arise and I´ll keep you informed.

    Regarding your questions:

    1. No, the server affected has low user connections

    2. After the iisreset the process id 21272 no longer exists, I assume it was related to IIS. If the problem occurs again, I´ll take note on that.

    3. Yes

    4. No

    5. I don´t know, I´ll monitor this

    Thank you for your help.

    Regards

    Thursday, September 25, 2014 12:34 PM
  • Eric, same problem again.

    'MSExchangeRpcProxyAppPool' gave event ID 5138 and then 5013 immediately.
    The same with 'MSExchangeSyncAppPool' and 'MSExchangeRpcProxyFrontEndAppPool'.

    Regards

    Thursday, September 25, 2014 8:30 PM
  • Hello,

    It sounds like there's no real impact currently. I found some data indicates that disabling the Microsoft Exchange Health Manager service may do help. This service is a controller process used to manage worker processes. There won’t be any impact for current environment if we disable this service.

    Thanks.


    Regards, Eric Zou

    Friday, September 26, 2014 7:25 AM
  • Obviously disabling the health manager service will stop generating the alert, but it does not resolve the problem. Furthermore, I will loose every monitoring alerts in this Exchange Server.

    I need to determine the real cause behind the application pool recycling.

    Thank you

    Regards

    Friday, September 26, 2014 1:39 PM
  • Hi Gonzalo,

    We could go through the following link:

    http://blogs.msdn.com/b/johan/archive/2007/05/16/common-reasons-why-your-application-pool-may-unexpectedly-recycle.aspx

    and check the related configuration.

    Moreover, if we'd like to determine the real cause, we may need to analyze more information such as the IIS log and http log. I suggest submitting a ticket to Microsoft IIS team to troubleshoot further.

    Thanks.


    Regards, Eric Zou

    • Marked as answer by Eric ZouZou Wednesday, October 8, 2014 8:12 AM
    Saturday, September 27, 2014 5:23 AM