Hi Michele,
I tried a cache size as high as 500000. It did not
seem to make a difference. As I had mentioned before,
all publishers/subscribers work on the same topic(
each message about 2KB - changed the program ). I
tried increasing the per-topic cache to 100k messages
and also the maxBytesCache to 100M . It did not help.
Thanks,
Siddhartha S
--- Michele Laghi <laghi at swissinfo.org> wrote:
Hi Siddhartha,
I tried the same as you did and got an overflow in
the cache queue. I
believe you got the same problem. The publishers are
faster than the
subscribers.
So I suggest you increase the maximum size of your
queue:
for example:
queue/callback/maxEntriesCache = 50000
queue/callback/maxEntries =
${queue/callback/maxEntriesCache}
Regards
Michele
Siddhartha S wrote:
Hi,
I am attetmpting to run the TestStress program
with 15 subscribers/publishers publishing to the
same
queue.
Many of the programs are exiting with the
following
error message:-
----------
Caught exception in publish #245 err
orCode=resource.exhaust,
message=[/home/sidharta/xmlBlaster/build.tmp/src/c/socket/XmlBlasterAccessUnparsed.c:406]
Waiting on response for 'publish()' with
requestId=248
timed out after blocking 60000 millis
----------
Each publisher attempts to publish 10000 messages.
The server is started as:-
java -server -Xms256M -Xmx512M -jar
lib/xmlBlaster.jar
What could be wrong here?
Thanks,
Siddhartha S
--
Michele Laghi
mailto:laghi at swissinfo.org
tel. +46 8 7492952 / mob. +46 70 4103964
http://edis.name
http://www.xmlBlaster.org