X server crashing after DB2 installation

Latest response

Hello,

I got an issue concerning a X server crash and "strange" behaviour (like backspace not erasing written stuff in terminal/console but rewriting letters backwards. e.g. hello world/ /dlrow olleh).

This behaviour first appeared on different machines where I installed DB2 v10.1 Enterprise. On machines where DB2 is not installed, these issues are not appearing. I am using RHEL 6.5.

I attached the Xorg.0.log.
Since I am a rather unexperienced Linux user I hope you can forgive me if a solution might be obvious.

Attachments

Responses

Hello again,

in the meantime two different IBM supporters told me that this wouldn't be an DB2 issue and has not been reported to them yet, so it seems to be an RHEL issue. Some of you got an idea what I could check to get rid of this issue?

Thanks

Silly question: when DB2 is installed, its installation changing any of your system-wide linker search orders (typically by modifying, contents of /etc/ld.so.conf or sometimes setting an LD_LIBRARY_PATH in a top order run control script) or your system's LOCALE? Both such changes can create unexpected behaviors.

Hi Christoph,

Let me reach out to some folks here who might be able to help with this one. I should have an answer for you soon.

Christoph, sorry for the delayed response here. It looks like the best choice of action here would be to open a support case. A coredump of the crashed process would be needed to further analyze the crash.

Hello Christoph,

I had worked on similar issue in past where disabling db2fmcd from startup was workaround identified and Upgrade of IBM DB2 to latest (version not known to me) release was confirmend by affected customer.

I am not sure if that applies here as well, If possible you may try it out.

To investigate further, We may need core of the process crashing (Xorg) along with list of installed rpms on affected system. Details should already be captured in /var/spool/abrt/ directory.
If disabling db2fcmd and/or using latest db2 does not help, I would suggest to open support case with those details.

If crash report is not captured in /var/spool/abrt/ directory and the problem is always reproducible, you may follow below steps to capture core..

1) Boot system in runlevel 3 and login to the console there.
2) Enable core dump
ulimit -c unlimited
3) Start GUI
startx
4) Reproduce issue
5) Core should be captured in current working directory from where startx command was issued.
ll core*
6) Generate sosreport after reproducing issue.
/usr/sbin/sosreport -a
7) Open support case with sosreport and core dump attached to it.

Regards,
Ashish

Can you confirm whether the upstart version installed is >= upstart-0.6.5-13.el6_5.3. If not can you update it and check. db2 installation creates /etc/init/db2fmcd.conf which has "console output". this updated upstart package resolves a related issue which should help

Close

Welcome! Check out the Getting Started with Red Hat page for quick tours and guides for common tasks.