[
https://issues.apache.org/jira/browse/HBASE-22648?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16891505#comment-16891505
]
Hudson commented on HBASE-22648:
--------------------------------
Results for branch branch-1
[build #968 on
builds.a.o|https://builds.apache.org/job/HBase%20Nightly/job/branch-1/968/]:
(x) *{color:red}-1 overall{color}*
----
details (if available):
(x) {color:red}-1 general checks{color}
-- For more information [see general
report|https://builds.apache.org/job/HBase%20Nightly/job/branch-1/968//General_Nightly_Build_Report/]
(x) {color:red}-1 jdk7 checks{color}
-- For more information [see jdk7
report|https://builds.apache.org/job/HBase%20Nightly/job/branch-1/968//JDK7_Nightly_Build_Report/]
(x) {color:red}-1 jdk8 hadoop2 checks{color}
-- Something went wrong running this stage, please [check relevant console
output|https://builds.apache.org/job/HBase%20Nightly/job/branch-1/968//console].
(/) {color:green}+1 source release artifact{color}
-- See build output for details.
> Snapshot TTL
> ------------
>
> Key: HBASE-22648
> URL: https://issues.apache.org/jira/browse/HBASE-22648
> Project: HBase
> Issue Type: New Feature
> Components: snapshots
> Affects Versions: 3.0.0
> Reporter: Andrew Purtell
> Assignee: Viraj Jasani
> Priority: Minor
> Fix For: 3.0.0, 1.5.0, 2.3.0
>
> Attachments: HBASE-22648-branch-1.patch, HBASE-22648-branch-1.patch,
> HBASE-22648-branch-2.patch, HBASE-22648-master-v2.patch,
> HBASE-22648-master-v3.patch, HBASE-22648-master-v4.patch,
> HBASE-22648-master-v5.patch, HBASE-22648-master-v6.patch,
> HBASE-22648-master-v8.patch, HBASE-22648-master.patch, Screen Shot 2019-07-10
> at 8.49.13 PM.png, Screen Shot 2019-07-10 at 8.52.30 PM.png, Screen Shot
> 2019-07-10 at 9.06.36 PM.png, Screen Shot 2019-07-16 at 11.06.03 AM.png
>
>
> Snapshots have a lifecycle that is independent from the table from which they
> are created. Although data in a table may be stored with TTL the data files
> containing them become frozen by the snapshot. Space consumed by expired
> cells will not be reclaimed by normal table housekeeping like compaction.
> While this is expected it can be inconvenient at scale. When many snapshots
> are under management and the data in various tables is expired by TTL some
> notion of optional TTL (and optional default TTL) for snapshots could be
> useful. It will help prevent the accumulation of junk files by automatically
> dropping the snapshot after the assigned TTL, making their data files
> eligible for cleaning. More comprehensive snapshot lifecycle management may
> be considered in the future but this one case is expected to be immediately
> useful given TTls on data are commonly applied for similar convenience.
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
