To track down the 'No meat found' issue we need the
1. PublishQos of your topic "ComponentHeartbeat" (as Michele mentioned
below)
2. If possible a log file with logLevel TRACE & CALL so we can follow
one of the "ComponentHeartbeat"
messages (how/when it arrives, how it is processed and finally delivered)
3. Please check if the hard disk is full
4. The sizes of $HOME/tmp/xmlBlaster_* (the four HSQLDB file sizes)
5. Please check in stderr logs for OutOfMemory exceptions
Your xmlBlaster is compiled with Sun-1.4.2_06 and running with
Blackdown-1.4.1-01,
i'll try this combination later ...
thanks
Marcel
Michele wrote:
Hi Marshall,
Marshall Shapiro wrote:
Also the publishQos of the messages would be interesting. (to get
them you could temporarly switch dump[socket] on).
Regards
Michele
dump[socket] seems to dump the whole message to the log. Many of the
messages are very large(5MB+). Is there a way to get the publishQos
information without the message content?
Just send us one/two typical publishQos and remove the rest if
possible. You could get them from the UpdateQos.toXml() and dump these
on disk or just take a cut&paste extract of the dump.
Regards
Michele
Marshall
