jackrabbit-oak-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jukka Zitting <jukka.zitt...@gmail.com>
Subject Re: Observation throughput
Date Fri, 21 Jun 2013 09:46:56 GMT
Hi,

On Friday, June 21, 2013, Michael D├╝rig wrote:
>
> AFAICS the call to DBCursor.hasNext seems to be the bottleneck here.


I'd look higher up the stack. The
AbstractNodeState.compareAgainstBaseState() method exists as an unoptimised
fallback, so having it show up in a performance bottleneck is a sign of a
missing implementation-specific optimised version of the comparison.
Figuring out how to avoid the fallback in this case should solve the issue.

BR,

Jukka Zitting

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message