|Title: ||How to diagnose a eloqdb6 hang|
|Author: ||Michael Marxmeier (email@example.com)|
How to diagnose a eloqdb6 hang
The eloqdb6 server process runs the Eloquence databases.
It should never hang. However strange things happen
and a small number of bugs have caused eloqdb6 to
hang in the past.
This article explains how to collect the details
the Eloquence support needs to understand and fix the problem.
A hang of the eloqdb6 process could be caused by an internal
deadlock situation. The thread list that could be obtained
from the eloqdb6 process provides information on the
internal thread status.
- What recent messages are in the eloqdb6 logfile?
- Was there any unusual system event recently?
Please review the system (HP-UX: /var/adm/syslog/syslog.log,
Linux: /var/log/messages) and provide any relevant messages.
- Do all processes that access the database hang
or only specific ones?
- Does the eloqdb6 process consumes CPU time?
- Please try to execute a "dbctl list thread"
If successful, please provide the output of this command.
- In case the "dbctl list thread" command hangs and
the HTTP status was enabled, please use a web browser to
obtain the thread status.
- In case the "dbctl list thread" command hangs,
please try to send a USR1 signal to the main eloqdb6
process (kill -USR1 process_id,
you need to be the eloqdb6 or root user).
The main eloqdb6 process is usually the first eloqdb6
process in a ps listing and has a parent process id 1.
This causes a dump of the internal thread status
to the eloqdb6 logfile.