[jira] [Comment Edited] (LUCENE-9827) Small segments are slower to merge due to stored fields since 8.7

["Robert Muir (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/LUCENE-9827?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17305741#comment-17305741 ] 

Robert Muir edited comment on LUCENE-9827 at 3/22/21, 2:37 AM:
---------------------------------------------------------------

if we wanted, in the future we could substitute something like {{avgDocsPerChunk = (numDocs - numDirtyDocs) / (numChunks - numDirtyChunks)}} instead of just using {{maxDocsPerChunk}} parameter as the threshold. It would allow us to more aggressively recompress small segments sometimes, adapting to different document sizes based upon the data we see.

But this adds some complexity (e.g. to testing) and I would prefer to do anything like that as a followup. By using avg estimate we also may end up still doing wasted work, vs. maxDocsPerChunk which is easier to reason about: gives us some guaranteed behavior.

edit: fix formula and clarify


was (Author: rcmuir):
if we wanted, in the future we could substitute something like {{avgDocsPerChunk = (numChunks - numDirtyChunks) / (numDocs - numDirtyDocs)}} instead of just using {{maxDocsPerChunk}} parameter as the threshold, if we want to more aggressively recompress small segments sometimes, adapting to different document sizes based upon the data we see.

But this adds some complexity (e.g. to testing) and I would prefer to do anything like that as a followup.

> Small segments are slower to merge due to stored fields since 8.7
> -----------------------------------------------------------------
>
>                 Key: LUCENE-9827
>                 URL: https://issues.apache.org/jira/browse/LUCENE-9827
>             Project: Lucene - Core
>          Issue Type: Bug
>            Reporter: Adrien Grand
>            Priority: Minor
>         Attachments: Indexer.java, log-and-lucene-9827.patch, merge-count-by-num-docs.png, merge-type-by-version.png, total-merge-time-by-num-docs-on-small-segments.png, total-merge-time-by-num-docs.png
>
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> [~dm] and [~dimitrisli] looked into an interesting case where indexing slowed down after upgrading to 8.7. After digging we identified that this was due to the merging of stored fields, which had become slower on average.
> This is due to changes to stored fields, which now have top-level blocks that are then split into sub-blocks and compressed using shared dictionaries (one dictionary per top-level block). As the top-level blocks are larger than they were before, segments are more likely to be considered "dirty" by the merging logic. Dirty segments are segments were 1% of the data or more consists of incomplete blocks. For large segments, the size of blocks doesn't really affect the dirtiness of segments: if you flush a segment that has 100 blocks or more, it will never be considered dirty as only the last block may be incomplete. But for small segments it does: for instance if your segment is only 10 blocks, it is very likely considered dirty given that the last block is always incomplete. And the fact that we increased the top-level block size means that segments that used to be considered clean might now be considered dirty.
> And indeed benchmarks reported that while large stored fields merges became slightly faster after upgrading to 8.7, the smaller merges actually became slower. See attached chart, which gives the total merge time as a function of the number of documents in the segment.
> I don't know how we can address this, this is a natural consequence of the larger block size, which is needed to achieve better compression ratios. But I wanted to open an issue about it in case someone has a bright idea how we could make things better.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org