On Mon, May 7, 2012 at 12:33 PM, Alfred von Campe <alf...@von-campe.com>wrote:
> On May 7, 2012, at 15:20, Christian Hammond wrote:
> That said, we do have a 1MB limit for diffs now (but that's not what
> you're hitting), as Review Board can get bogged down if too many people are
> processing large diffs at once. It's not configurable yet, but in general,
> very large diffs can and should generally be avoided. They're quite hard to
> review, and usually mean that there's some auto-generated code or something
> in them that could be split out for the sake of review purposes.
> Thanks for the quick response (as always). We were able to work around
> this by removing the project file from the patch file and that reduced the
> size sufficiently that it accepted the review.
> On a somewhat related issue, the performance of our ReviewBoard server (or
> lack thereof) has been a frequent source of complaints recently so I am
> probably going to set up a new server. What is the biggest factor
> affecting ReviewBoard performance? CPU speed, CPU count, available memory,
> disk speed, database, etc.? I'd like to configure a new server (or
> possibly a VM) appropriately to get better performance. What should I look
> out for?
Memory and how much of it that memcached can use is crucial. We cache a
*lot*, since fetching files from the repository, patching them, and
generating diffs is all very expensive. So the more that memcached can hold
at once, the faster things will feel all around.
After that, you'll get some gains from database optimization and from
increased CPU performance (for the diff generation).
How big is your userbase, and what are your current specs? I know of
servers with thousands of users that stand up under constant use. Usually
it's just a configuration issue, or lack of memory for caching, that causes
the most problems.
Christian Hammond - chip...@chipx86.com
Review Board - http://www.reviewboard.org
VMware, Inc. - http://www.vmware.com
Want to help the Review Board project? Donate today at
Happy user? Let us know at http://www.reviewboard.org/users/
To unsubscribe from this group, send email to
For more options, visit this group at