[
https://issues.apache.org/jira/browse/LUCENE-10448?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17510491#comment-17510491
]
Adrien Grand commented on LUCENE-10448:
---------------------------------------
bq. it is guaranteed that there will be no big chunks written to disk in
theory, what am I missing?
This is where I'm confused too. Based on the discussion and [~vigyas]'s PR, it
feels to me like the problem stems from writing big chunks after a period of
inactivity, which is why Vigya's PR that splits big chunks into smaller ones
helps honor the instant rate.
So either Lucene writes big chunks at times, and then I'd rather look into
these large writes to understand if we should be doing things differently
instead of just fixing the problem at the rate limiter level which could delay
the garbage collection of these large byte[] arrays.
Or Lucene does not write big chunks of data, but then I don't understand what
is the bug that we are discussing in this issue. Sure, some writes might not
honor the instant rate limit at times, but since these are small chunks we
don't care?
> MergeRateLimiter doesn't always limit instant rate.
> ---------------------------------------------------
>
> Key: LUCENE-10448
> URL: https://issues.apache.org/jira/browse/LUCENE-10448
> Project: Lucene - Core
> Issue Type: Bug
> Components: core/other
> Affects Versions: 8.11.1
> Reporter: kkewwei
> Priority: Major
> Time Spent: 2.5h
> Remaining Estimate: 0h
>
> We can see the code in *MergeRateLimiter*:
> {code:java}
> private long maybePause(long bytes, long curNS) throws
> MergePolicy.MergeAbortedException {
>
> double rate = mbPerSec;
> double secondsToPause = (bytes / 1024. / 1024.) / rate;
> long targetNS = lastNS + (long) (1000000000 * secondsToPause);
> long curPauseNS = targetNS - curNS;
> // We don't bother with thread pausing if the pause is smaller than 2
> msec.
> if (curPauseNS <= MIN_PAUSE_NS) {
> // Set to curNS, not targetNS, to enforce the instant rate, not
> // the "averaged over all history" rate:
> lastNS = curNS;
> return -1;
> }
> ......
> }
> {code}
> If a Segment is been merged, *maybePause* is called in 7:00, lastNS=7:00,
> then the *maybePause* is called in 7:05 again, so the value of
> *targetNS=lastNS + (long) (1000000000 * secondsToPause)* must be smaller than
> *curNS*, no matter how big the bytes is, we will return -1 and ignore to
> pause.
> I count the total times(callTimes) calling *maybePause* and ignored pause
> times(ignorePauseTimes) and detail ignored bytes(detailBytes):
> {code:java}
> [2022-03-02T15:16:51,972][DEBUG][o.e.i.e.I.EngineMergeScheduler] [node1]
> [index1][21] merge segment [_4h] done: took [26.8s], [123.6 MB], [61,219
> docs], [0s stopped], [24.4s throttled], [242.5 MB written], [11.2 MB/sec
> throttle], [callTimes=857], [ignorePauseTimes=25], [detailBytes(mb) =
> [0.28899956, 0.28140354, 0.28015518, 0.27990818, 0.2801447, 0.27991104,
> 0.27990723, 0.27990913, 0.2799101, 0.28010082, 0.2799921, 0.2799673,
> 0.28144264, 0.27991295, 0.27990818, 0.27993107, 0.2799387, 0.27998447,
> 0.28002167, 0.27992058, 0.27998066, 0.28098202, 0.28125, 0.28125, 0.28125]]
> {code}
> There are 857 times calling *maybePause*, including 25 times which is ignored
> to pause, we can see that the ignored detail bytes (such as 0.28125mb) are
> not small.
> As long as the interval between two *maybePause* calls is relatively long,
> the pause action that should be executed will not be executed.
>
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
