• 1 reply
  • Latest Post - ‏2014-06-02T19:43:08Z by FrankBrandt
Awais Sheikh
Awais Sheikh
4 Posts

Pinned topic Warnings in TDS log file

‏2014-05-29T18:40:15Z |
Hi All,

I am using TDS as our user repository behind WebSphere Portal. There is a warning that I usually see in ibmslapd log file and i.e.    

"GLPSRV204W The server has temporarily suspended reading client requests from the network 13530 times. There are 1 of 15 worker threads attempting to write results." 

Also, there is another informational message that I think is associated with it:

"GLPSRV205I The server has resumed reading client requests from the network."
After doing a little research I increased a value of ibm-slapdDbConnections from 15 to 60 but still the warning persist. This warning appears even at a time when the server is having minimum traffic. Although these warning does not effect the system, yet it makes me curious with its existence.

I am looking for the reasons behind such warnings as well as any sort of remediation that can vanish it permanently. Our ldap server is tuned so far, but I will be glad to know if there are any specific tuning parameters need to be changed.



  • FrankBrandt
    18 Posts

    Re: Warnings in TDS log file



    we have the same message from time to time. We have ibm-slapdDbConnections set to 40. I think someone back in the TDS beta said not to increase the value higher as this can result in database locks (at least from what I can remember). If you are a DB2 guru maybe 60 is ok :)

    In general I have no direct solution to resolve the problem. What we did is enable the audit log for searches on production for a week and got some gigabytes of log. There are a bunch of Perl scripts from IBM to search for slow queries and unindexed attributes in the audit log. You can also check the audit.log if the Portal sends some bad requests (from performance perspective of the LDAP server - We had some "nice" application connected to our TDS doing the following: Do a request to get all groups relevant for the application. Then foreach member do 1 seperate ldap search and query the person. This resulted in a huge number of queries in a short amount of time and resulted in the error above. )

    Good luck finding the problem