markrmiller commented on a change in pull request #230:
URL: https://github.com/apache/solr/pull/230#discussion_r706265995
##########
File path: solr/solr-ref-guide/src/caches-warming.adoc
##########
@@ -79,6 +79,9 @@ Reasonable values, depending on the query volume and
patterns, may lie somewhere
The `maxRamMB` attribute limits the maximum amount of memory a cache may
consume.
When both `size` and `maxRamMB` limits are specified the `maxRamMB` limit will
take precedence and the `size` limit will be ignored.
+The `async` attribute determines whether the cache stores direct results
(`async=false`, the default) or whether it will store indirect references to
the computation (`async=true`). Enabling `async` will use more slightly more
memory per cache entry, but may result in reduced CPU cycles generating results
or doing garbage collection. The most noticable improvements will be seen when
there are many queries racing to compute the same results before they are
inserted into the cache. In some use cases, increasing autowarming may be a
better alternative to enabling an async cache. Additionally, an async cache
will not prevent a data race for time-limited queries, as those may return
differing sets of partial results.
Review comment:
Nothing urgent here, but personally I'd like it to also be a little more
clear that this issue can help solve the problem in the description of the
JIRA. It's somewhat referenced with the mention of autowarming, but with no
history of that issue, I'd have no idea that using this would help prevent that
problem or why autowarming was somehow another approach to the async option.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
