During the tests I've run with Raven on Azure I've noticed something interesting - one of the servers is consistently using 20% of the CPU time, while the other is using only 2%. This is when idle; heck the server in question, while being a master, is not directly referenced by any apps and I guess it only gets round-robin read traffic, which is VERY low. Here's a New Relic screenshot: http://imgur.com/NhUSsxwHow can I even begin to diagnose this? I certainly don't want to rebuild the VM, as I've done that too many times :D
--
You received this message because you are subscribed to the Google Groups "ravendb" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ravendb+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
I remember a similar behavior. The source of the problem turned out to be conflict documents we weren't handling, iirc. Adding a server side conflict resolver made it go away.
Any conflicted docs?
--