ODBC Connector Issue

kumar66 · Post by **kumar66** » Fri Jan 06, 2012 5:46 am

Hi All,

when I tried to connect Oracle through ODBC Connector Stage and hit Test button , I am getting the following error:

An exception occured while trying to send the request to the handler:The agent at ist:31531 is not avaliable.

I did the following :

stopped the ASBNode agent
Stopped the Logging Agent
Started the ASBNode Agent and
Staretd the Logging Agent

But still encoutering the same error.

Please Advise
Kumar66

qt_ky · Post by **qt_ky** » Fri Jan 06, 2012 8:48 am

I've been having that also and have a PMR open about it. I can restart the agent using the NodeAgents.sh stop and start options, and it's good for a while but dies again within a few hours to a day. We have found javacore dump files in the ASBNode directory that indicate java out of memory errors. We have added a memory setting to the Agent.sh script that did not help. We have added an AIX-specific setting to the same script that did not help. We're on 8.5 and they have suggested a possible memory leak and the next step would be to apply fix pack 1. If you search on the above errors or key words on the IBM Support Portal, you will find tech notes showing the script changes and suggesting they are fixed in 8.1 FPx or even in 8.5. I don't know the root cause just yet, but when the agent is broken, we get the same error on any connector stage and all the ISD web service jobs fail too.

kumar66 · Post by **kumar66** » Mon Jan 09, 2012 9:37 am

Hi qt_ky,

Thanks for your reply. After the reference from IBM Support portal , I added the entries to in Agent.sh file.

But now I am getting the follwoing error when I click the test button in ODBC connector stage:

" An exception occured while trying to send the request to the handler: Connection refused to host :107.21.159.192 ; nested exception is:java.net.ConnectException : Connection timed out"

Please Advise.
Thanks ,
kumar66

kumar66 · Post by **kumar66** » Tue Jan 10, 2012 12:49 am

Hi All,

Any suggestion on this?

Please Advice.

Regards,
Kumar66

chulett · Post by **chulett** » Tue Jan 10, 2012 7:39 am

Contact your official support provider. That and probably undo whatever change you just made.

kumar66 · Post by **kumar66** » Thu Jan 12, 2012 12:41 am

Hi Chulett,

Just for an info, the datastage server is installed on Amazon Cloud.

Thanks,
Kumar66

qt_ky · Post by **qt_ky** » Sun Jan 15, 2012 7:29 pm

In my case, java would crash with out of memory errors immediately upon restarting the node agent. The error was found in a java core dump file and NOT written to the screen at all. Support had me make a few Agent.sh changes and for the memory leak permanent fix, apply 8.5 Fix Pack 1.

Madhumitha_Raghunathan · Tue Aug 21, 2012 9:05 am

Hi Eric,

We have IIS 8.7 and we are facing the same issue. The Node Agent keeps dying and all the connectors go down and we need to restart the NodeAgent and DSEngine.

On your PMR for Memory Leak did you get any resolution for this in terms of configuration changes?

Thanks,
Madhumitha