none
ProjectPublishEnqueueFailure with psi call RRS feed

  • Question

  • Hi:

    We are currently running Sharepoint 2007, Project Server 2007 with the latest service pack. 

    I am using PSI call to check out my project, update the data, check the project back in and publish the project.  But sometimes I have encounter error message.     System.Web.Services.Protocols.SoapException: ProjectServerError(s) LastError=ProjectPublishEnqueueFailure Instructions: Pass this into PSClientError constructor to access all error information
    at Microsoft.Office.Project.Server.WebService.Project.

    My data looks like it updated correctly.  The project will not be checkin whenever this error occurs.   Not sure if anyone has encounter a similar issue.  If I run, the process again.  It always finish successfully. I am not sure if the queue is getting messed up. 

    Thanks,

    kkmick

     

     

    Monday, October 3, 2011 8:57 PM

Answers

  • Hi there,

    If you cannot reproduce the problem for every PSI call, maybe your queue is the problem; Any jobs getting stuck before adding your one.

    Finally we decided to write a script to restart queue service every night and avoiding queue stuck.

    Works fine since 2 years now...

    Hope this helps...but not sure !

    • Marked as answer by kkmick Tuesday, October 4, 2011 12:24 PM
    Monday, October 3, 2011 9:44 PM
  • Hi,
    I have previously had similar issues with PSI development in 2007, I logged a case with Microsoft about this specifically as I was able to consistently cause queue failure trying to update and publish projects when calling PSI from a web service. (on SP2 etc)
    I have dug this up from my email, is this similar to the error you get in the ULS:
    Unhandled Exception: System.Web.Services.Protocols.SoapException:
    System.Web.Services.Protocols.SoapException: Server was unable to process
    request. ---> System.Xml.XmlException: Root element is missing.

    If this is similar then I cannot unfortunately offer a great solution, In my case we had to basically handle the above error and retry the job when it failed. Specifically however the consistent 20% failure rate was only during times of high server load, so scaling out our servers and actually slowing down our code solved 99% of the failures..

     

    HTH,


    Martin Laukkanen (Project Server Blog - www.nearbaseline.com/blog)
    • Marked as answer by kkmick Tuesday, October 4, 2011 12:24 PM
    Tuesday, October 4, 2011 12:26 AM

All replies

  • Hi there,

    If you cannot reproduce the problem for every PSI call, maybe your queue is the problem; Any jobs getting stuck before adding your one.

    Finally we decided to write a script to restart queue service every night and avoiding queue stuck.

    Works fine since 2 years now...

    Hope this helps...but not sure !

    • Marked as answer by kkmick Tuesday, October 4, 2011 12:24 PM
    Monday, October 3, 2011 9:44 PM
  • Hi,
    I have previously had similar issues with PSI development in 2007, I logged a case with Microsoft about this specifically as I was able to consistently cause queue failure trying to update and publish projects when calling PSI from a web service. (on SP2 etc)
    I have dug this up from my email, is this similar to the error you get in the ULS:
    Unhandled Exception: System.Web.Services.Protocols.SoapException:
    System.Web.Services.Protocols.SoapException: Server was unable to process
    request. ---> System.Xml.XmlException: Root element is missing.

    If this is similar then I cannot unfortunately offer a great solution, In my case we had to basically handle the above error and retry the job when it failed. Specifically however the consistent 20% failure rate was only during times of high server load, so scaling out our servers and actually slowing down our code solved 99% of the failures..

     

    HTH,


    Martin Laukkanen (Project Server Blog - www.nearbaseline.com/blog)
    • Marked as answer by kkmick Tuesday, October 4, 2011 12:24 PM
    Tuesday, October 4, 2011 12:26 AM
  • Thanks Martin and Jeff for your answer.  Jeff, I actually have tried to slow down.  It seems to work a bit better than before.   I will definitely keep Martin's idea about  writing the script to restart queue service on nightly basis.  

    We are moving to Project Server 2010.  I am trying to sort things out before the migration.  I am hoping the new project server 2010 psi call will fix will the issue.   I  just want to make sure it was not my psi code that I create cause that particular issue. 

     

    Thanks for the suggestions. 

    kkmick

     

     

     


    • Edited by kkmick Tuesday, October 4, 2011 12:26 PM
    Tuesday, October 4, 2011 12:24 PM