[ https://issues.apache.org/jira/browse/LUCENE-7976?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16427982#comment-16427982 ]
Marc Morissette commented on LUCENE-7976: ----------------------------------------- [~erickerickson] Thanks for tackling this. Regarding singleton merges: if I read your code correctly and am right about how Lucene works, I think that, on a large enough collection, your patch could generate ~50% more reads/writes when re-indexing the whole collection: * I think new documents are typically flushed once and merged 2-3 times before ending up in a large segment. * With a 20% delete threshold, old documents would, on average, be singleton merged 4 times before being expunged vs only one merge at a 50% delete threshold. In Latex notation: {code:java} 20% deleted docs threshold: \sum_{n=1}^\infnty (1 - 0.2)^n = (1 / (1 - (1 - 0.2))) - 1 = 4 50% deleted docs threshold: \sum_{n=1}^\infnty (1 - 0.5)^n = (1 / (1 - (1 - 0.5))) - 1 = 1{code} On the odd chance that my math bears any resemblance to reality, I would suggest that you disable singleton merges when the short term deletion rate of a segment is above a certain threshold (say 0.5% per hour). This should prevent performance degradations during heavy re-indexation while maintaining the desired behaviour on seldom updated indexes. > Make TieredMergePolicy respect maxSegmentSizeMB and allow singleton merges of > very large segments > ------------------------------------------------------------------------------------------------- > > Key: LUCENE-7976 > URL: https://issues.apache.org/jira/browse/LUCENE-7976 > Project: Lucene - Core > Issue Type: Improvement > Reporter: Erick Erickson > Assignee: Erick Erickson > Priority: Major > Attachments: LUCENE-7976.patch, LUCENE-7976.patch > > > We're seeing situations "in the wild" where there are very large indexes (on > disk) handled quite easily in a single Lucene index. This is particularly > true as features like docValues move data into MMapDirectory space. The > current TMP algorithm allows on the order of 50% deleted documents as per a > dev list conversation with Mike McCandless (and his blog here: > https://www.elastic.co/blog/lucenes-handling-of-deleted-documents). > Especially in the current era of very large indexes in aggregate, (think many > TB) solutions like "you need to distribute your collection over more shards" > become very costly. Additionally, the tempting "optimize" button exacerbates > the issue since once you form, say, a 100G segment (by > optimizing/forceMerging) it is not eligible for merging until 97.5G of the > docs in it are deleted (current default 5G max segment size). > The proposal here would be to add a new parameter to TMP, something like > <maxAllowedPctDeletedInBigSegments> (no, that's not serious name, suggestions > welcome) which would default to 100 (or the same behavior we have now). > So if I set this parameter to, say, 20%, and the max segment size stays at > 5G, the following would happen when segments were selected for merging: > > any segment with > 20% deleted documents would be merged or rewritten NO > > MATTER HOW LARGE. There are two cases, > >> the segment has < 5G "live" docs. In that case it would be merged with > >> smaller segments to bring the resulting segment up to 5G. If no smaller > >> segments exist, it would just be rewritten > >> The segment has > 5G "live" docs (the result of a forceMerge or optimize). > >> It would be rewritten into a single segment removing all deleted docs no > >> matter how big it is to start. The 100G example above would be rewritten > >> to an 80G segment for instance. > Of course this would lead to potentially much more I/O which is why the > default would be the same behavior we see now. As it stands now, though, > there's no way to recover from an optimize/forceMerge except to re-index from > scratch. We routinely see 200G-300G Lucene indexes at this point "in the > wild" with 10s of shards replicated 3 or more times. And that doesn't even > include having these over HDFS. > Alternatives welcome! Something like the above seems minimally invasive. A > new merge policy is certainly an alternative. -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org