Hi Guys, The subject of this question is the error which I am receiving when running a scheduled job also this is an error which we get at a random intervals and the frequency would be once or twice in a month. I have been through several posts and didn't concluded anything out of it. At the time when we receive this error the memory usage of the database was normal with free resources available and also the network was stable without issues. Our Sql database is in cluster mode with named instance. Can anyone share some information on this, I know this is not the right forum with the error message I have posted but am sure some one might have face the same problem when running some ETL jobs. Source and target both are sql server and the error message is generated in the validation phase of the connection manager connecting to target database. SSIS 2005 SP2 64 bit environmentRaju

Hi Guys, Can any one shed some light on this issue. We came across the same issue again today morning and as a workaround we had manually kill all the process which starts with dtexec.exe and ran again the same job which completed successfully without issues. This work around process has been followed from long time now. Just curious to know the actual cause of this issue.Raju

Hello Raju, My wild guess here is that your clustered environment hiccuped for a short period of time, thus became intermittently unavailable and/or perhaps wanted to flip nodes. You may want to start examining the cluster event logs. Please let us know what you find. If you have dedicated server administrators please ask her/him for help.Arthur My Blog

Hi Arthur. Yes you may be right. I have asked our INFRASTRUCTURE team to check these logs they will be getting back in a day. Will update the post once received the results.Raju

Hi Arthur, Bad Luck nothing happened from cluster point of view and also the network was stable during the failure times.Raju

Hi Raju, a negative result is also a result, I still think there was an unavailability issue, but no particular proof is available at this time. There are a couple of things we can do: 1) If not yet done, ask your DBA to send an alert to you in case the job fails again (this way you will have an accurate time of the failure event); The same time the job fails try to connect to the target and ask your dba to check SQL Server logs (may be the database was in a special mode); 2) Try to see if there is a correlation between this failure and the entries in the Windows Event Logs; and 3) Make sure your Windows or SQL server does not need any updates. Arthur My Blog

Hi Arthur, I do get an alert when something fails and I use to check the database with the same I can connect with out issues. Also when i see the error logs of sql server it has the below error: Login failed for user 'XXXX'. [CLIENT: 1.2.3.4] Error: 18456, Severity: 14, State: 8. This error is purely related to password mismatch, but the password was never changed and when the ETL is manually run for the second time with same login credentials it use to run with out isse and also there are no entires in the windows event logs which i can relate to this error. The error which is in the subject of this thread is captured by errordescription variable inside the ETL. Windows Server : Windows Server 2003 R2 Enterprise X64 Edition Sql Server : SQl Server 2005 Service pack2 Can you suggest what needs to be updated which will over come the actual issue. Raju

Try using a different account for your job which does not have the failures you mentioned.Arthur My Blog

From the Past 2 days this issue has dissappeared. I beleive this has got more to do with DBNETLIB driver. Raju

Can you elaborate more please on why do you think so?Arthur My Blog

I have enabled on Tracing on Sql client on ETL machine but from then I didn't got this error am waiting for the error to occur to check the client driver logs. Raju

Guys, There is still no luck in finding the root cause of this error. These errors started coming once and now it is happening atleast once in a week and the same workaround of rerunning the ETL's would solve without changing anything.Raju