Large sized files ODBC_GRE_*.log getting created in Analytics/bin folder

[shradha]

Large ODBC_GRE files are getting generated in our analytics folder filling us all the space and getting server to 0 bytes. They size upto 50 GB(Or may be more if space is left on server)

Could not find much how and what are these files. Anyone has any idea? Please help.

we have already disabled creation of core dump files by setting <env_var name="CCL_HWE_ABORT" value="0"/>

[MFGF]

Large ODBC_GRE files are getting generated in our analytics folder filling us all the space and getting server to 0 bytes. They size upto 50 GB(Or may be more if space is left on server)

Could not find much how and what are these files. Anyone has any idea? Please help.

we have already disabled creation of core dump files by setting <env_var name="CCL_HWE_ABORT" value="0"/>

Hi,

What data source(s) and driver(s) are you using? Are you seeing frequent report crashes? It sounds like you have a deeper issue than just the disk filling up. Were you getting core dump files before you disabled their creation? If so, did you look into what was causing the issue?

MF.

[shradha]

Hi MFGF,

Thanks for your reply.

We are using multiple data sources and drivers as we are working on multiple implementations. We use both CQM/DQM mode, and both ODBC/JDBC data connections.

This work has been going on for months but we have started facing this issue in last couple of days. And it's very frequent. The core dump files disabling was done months ago and system was working fine. We had issue of dump file generation for some reports which used multiple table unions, so we disabled it.

We have not experienced report crashes until the server runs out of space. The system is being used by multiple users from multiple locations.

We can see the same file from months , but they are not as big (just 2-10 KBs) as they are being generated in last couple of days. In these files (small ones), all we can see is the connection detail to the db server. We use optim as the db.

[MFGF]

Hi MFGF,

Thanks for your reply.

We are using multiple data sources and drivers as we are working on multiple implementations. We use both CQM/DQM mode, and both ODBC/JDBC data connections.

This work has been going on for months but we have started facing this issue in last couple of days. And it's very frequent. The core dump files disabling was done months ago and system was working fine. We had issue of dump file generation for some reports which used multiple table unions, so we disabled it.

We have not experienced report crashes until the server runs out of space. The system is being used by multiple users from multiple locations.

We can see the same file from months , but they are not as big (just 2-10 KBs) as they are being generated in last couple of days. In these files (small ones), all we can see is the connection detail to the db server. We use optim as the db.

Hi,

If core dumps are being generated, it's an indication that processes are crashing, which really doesn't tally with the message about the system working fine. I'd be looking into what the core dumps are telling you, and fix that issue if I were you.

For the ODBC log files, this may or may not be related to the same processes crashing, but because you've inhibited core dumps you're not going to be able to tell. You can also enable ODBC Trace to see if it helps you troubleshoot the issue - details of how to are here:

https://www.ibm.com/support/pages/enabling-odbc-trace-cognos-analytics

Good luck!

MF.