Adar Dembo has posted comments on this change.

Change subject: process_memory: go back to non-incremental tracking
......................................................................


Patch Set 1:

(3 comments)

http://gerrit.cloudera.org:8080/#/c/6915/1//COMMIT_MSG
Commit Message:

PS1, Line 34: noticeably better
> takes a long time to run, so I didn't wait for the whole thing to finish :)
Thanks. It looks like the difference between the two implementations is about 
1-2% CPU time. Am I reading the graph correctly?


http://gerrit.cloudera.org:8080/#/c/6915/1/src/kudu/util/process_memory.cc
File src/kudu/util/process_memory.cc:

Line 191:   const int64_t kReadIntervalMicros = 50000;
> figured that we would have a hard time using more than a few MB in 50ms, an
I thought we could consume more than that, even in such a short period of time. 
Like, ballpark 1M inserts/s, so 50K for every 50ms. With a 1K row size, that's 
50 MB.

Do you think that's achievable?


PS1, Line 198: GetMonoTimeMicros()
> I wanted to re-fetch time, so that we guarantee the sleep _between_ calls i
Makes sense. Could you add a comment explaining this, so no one "optimizes 
away" the second call?


-- 
To view, visit http://gerrit.cloudera.org:8080/6915
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I8823028de3ea260f1450d9bf34af2dc5a794b206
Gerrit-PatchSet: 1
Gerrit-Project: kudu
Gerrit-Branch: master
Gerrit-Owner: Todd Lipcon <[email protected]>
Gerrit-Reviewer: Adar Dembo <[email protected]>
Gerrit-Reviewer: Kudu Jenkins
Gerrit-Reviewer: Todd Lipcon <[email protected]>
Gerrit-HasComments: Yes

Reply via email to