[ https://issues.apache.org/jira/browse/CASSANDRA-7066?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14651571#comment-14651571 ]
Benedict commented on CASSANDRA-7066: ------------------------------------- bq. Will this cause problems if the alpha goes out before this ticket is committed? We should be fine, since we're changing on-wire serialization between alpha and beta also, so I guess we're OK with some upgrade problems between the two (we don't really want or expect anyone to use alpha except for dev, I would guess). [~jbellis]? bq. However all classes implementing MetadataComponent implement them so I left them with a comment. Also, without it it seems a scrub test was failing (unsure if related though). I guess it's possible we compare metadata at the end of a scrub to confirm it's equivalent. It's somewhat ugly that we will treat all compaction metadata as equal, but probably not worth worrying about here. Otherwise SGTM, I'll review it shortly. > Simplify (and unify) cleanup of compaction leftovers > ---------------------------------------------------- > > Key: CASSANDRA-7066 > URL: https://issues.apache.org/jira/browse/CASSANDRA-7066 > Project: Cassandra > Issue Type: Improvement > Components: Core > Reporter: Benedict > Assignee: Stefania > Priority: Minor > Labels: benedict-to-commit, compaction > Fix For: 3.0 alpha 1 > > Attachments: 7066.txt > > > Currently we manage a list of in-progress compactions in a system table, > which we use to cleanup incomplete compactions when we're done. The problem > with this is that 1) it's a bit clunky (and leaves us in positions where we > can unnecessarily cleanup completed files, or conversely not cleanup files > that have been superceded); and 2) it's only used for a regular compaction - > no other compaction types are guarded in the same way, so can result in > duplication if we fail before deleting the replacements. > I'd like to see each sstable store in its metadata its direct ancestors, and > on startup we simply delete any sstables that occur in the union of all > ancestor sets. This way as soon as we finish writing we're capable of > cleaning up any leftovers, so we never get duplication. It's also much easier > to reason about. -- This message was sent by Atlassian JIRA (v6.3.4#6332)