I've got two problems with my Navigator installation, and I don't know if they are related.
1. Navigator is only collecting a very few sessions.
For example, if I look at today's session report, it will tell me that there were 9 sessions. I dived into the postgresql database and confirmed that the report is genuinely telling me what it thinks is in the database.
DCT_vvslm0007=# select session_id from sessions where started > '2015-03-20'; session_id
----------------
2015/03/20-64
2015/03/20-65
2015/03/20-66
2015/03/20-67
2015/03/20-69
2015/03/20-68
2015/03/20-70
2015/03/20-71
2015/03/20-181
(9 rows)
But if I connect to the cell manager that I'm reporting on, there are hundreds of sessions, including backup sessions between 20-71 and 20-181.
Last weekend (when it was first installed) it collected a reasonable number of sessions but every day since then it has been very sparse, only collecting a few percent of the actual sessions.
Is there a way of forcing a re-fetch of a session?
2. /opt/dpa-ext/logs/agent_core.log is rapidly filling up with lines like this.
15-03-20 15:32:48 WARN () SessionReportsManipulation - Cannot execute omnidb - no access to binary
But I can't find on which cell manager it can't access omnidb. There are symlinks from /opt/omni/lbin/omnidb to /omni/omni/bin/omnidb on both cell managers. For good measure I even symlinked it to /usr/bin/omnidb in case it was being invoked by name.
I even installed the OMNICC component on the Navigator server just to see if it was called omnidb locally (it's not).
Could these problems be related somehow?