You are viewing a plain text version of this content. The canonical link for it is here.
Posted to notifications@couchdb.apache.org by gi...@git.apache.org on 2017/06/03 09:28:52 UTC
[GitHub] janl commented on issue #541: Allow configuration of max doc IDs for optimised code path in changes filter
janl commented on issue #541: Allow configuration of max doc IDs for optimised code path in changes filter
URL: https://github.com/apache/couchdb/pull/541#issuecomment-305963713
Solid PR!
The only question i have is why there is a threshold for the optimised path? The way I understand this there are two code paths, one is faster, but can?t deal with a large number of doc ids, the other is slower, but can deal with an unbounded number of doc ids. The question then is: when does the fast path turn into the slow path (or even-slower-than-the-slow-path) and why is the limit 100 now? What trade-off is being made?
I?d like to understand this if only to make sure that we add a comment in the .ini file that explains, hypothetically: ?you can set that to whatever, but for every 1k doc ids allowed, you?re using an extra 2Mb of memory per concurrent doc_id `_changes` request?, or whatever it ends up being, so people understand the trade-offs.
`git blame` tells me this code is originally @davisp?s, so ping :)
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
With regards,
Apache Git Services