On 31 Mar 2010, at 14:02, Alan Conway wrote:
On 03/30/2010 05:38 PM, Andrew Wright wrote:
Hi all,
I've recently run some tests to try and see how much overhead
clustering
brings. In short - I saw roughly a 50% reduction in message
throughput
when clients ran against a 2 node cluster vs a standalone broker. I'd
definitely expect some reduction, but perhaps not as much as that.
Does
that sound about right or would you expect it to be somewhat less if
appropriately configured?
I get about a 25% reduction running against a 4 node cluster.
Try --worker-threads=4.
I should point out that the standalone broker message throughput was
very good! So even a 50% reduction still leaves quite a bit of
speed in
absolute terms.
Cheers,
Andrew
Few more details:
- 200 byte msgs, sent/rcvd as fast as possible by clients
- servers used a single NIC, I'm going to try and re-run tests with a
dedicated NIC for AIS
That should help. You might try running perftest as it has ack
settings etc. set up for speed by default, see if that gives a
result different from you own clients.
Thanks for that Alan - 25% would certainly be more in line with what I
was expecting. I'll take a look at perftest, see if there's any tips
there.
Cheers,
Andrew
---------------------------------------------------------------------
Apache Qpid - AMQP Messaging Implementation
Project: http://qpid.apache.org
Use/Interact: mailto:[email protected]
