[
https://issues.apache.org/jira/browse/CASSANDRA-8038?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14161565#comment-14161565
]
Takenori Sato commented on CASSANDRA-8038:
------------------------------------------
Here's the end to end test result.
1. create 1M columns
2. delete all except for the first and the last
3. run slice query for the row
[RESULT]
ignore_tombstones_for_read_repair = false
[default@MyKeyspace] get TombstoneTest['test'];
=> (name=30, value=61676530, timestamp=1412649368159000)
=> (name=393939393939, value=616765393939393939, timestamp=1412649484332005)
Returned 2 results.
Elapsed time: 1563 msec(s).
===== FULL GC BEFORE =====
2014-10-07T15:54:33.177+0900: [Full GC2014-10-07T15:54:33.177+0900: [CMS:
3625K->14592K(786432K), 0.1062770 secs] 68746K->14592K(1022400K), [CMS Perm :
18501K->18497K(29076K)], 0.1064300 secs] [Times: user=0.10 sys=0.01, real=0.11
secs]
===== START =====
2014-10-07T15:54:41.276+0900: [GC2014-10-07T15:54:41.276+0900: [ParNew:
209792K->26176K(235968K), 0.4913860 secs] 224384K->190057K(1022400K), 0.4914810
secs] [Times: user=1.59 sys=0.13, real=0.49 secs]
INFO 06:54:41 ParNew GC in 492ms. CMS Old Gen: 14942488 -> 167814744; Par
Eden Space: 214827008 -> 0; Par Survivor Space: 0 -> 26804224
WARN 06:54:41 Read 2 live and 999998 tombstoned cells in
MyKeyspace.TombstoneTest (see tombstone_warn_threshold). 1000000 columns was
requested, slices=[-], delInfo={deletedAt=-9223372036854775808,
localDeletion=2147483647}
===== FULL GC AFTER =====
2014-10-07T15:54:52.382+0900: [Full GC2014-10-07T15:54:52.382+0900: [CMS:
163881K->13399K(786432K), 0.1719320 secs] 229431K->13399K(1022400K), [CMS Perm
: 18756K->18755K(30832K)], 0.1720680 secs] [Times: user=0.17 sys=0.00,
real=0.17 secs]
ignore_tombstones_for_read_repair = true
[default@MyKeyspace] get TombstoneTest['test'];
=> (name=30, value=61676530, timestamp=1412649368159000)
=> (name=393939393939, value=616765393939393939, timestamp=1412649484332005)
Returned 2 results.
Elapsed time: 1042 msec(s).
===== FULL GC BEFORE =====
2014-10-07T15:57:29.419+0900: [Full GC2014-10-07T15:57:29.419+0900: [CMS:
3637K->13829K(786432K), 0.1016750 secs] 91548K->13829K(1022400K), [CMS Perm :
17909K->17904K(29796K)], 0.1018380 secs] [Times: user=0.10 sys=0.01, real=0.10
secs]
===== START =====
2014-10-07T15:57:45.911+0900: [GC2014-10-07T15:57:45.911+0900: [ParNew:
209792K->11510K(235968K), 0.0095760 secs] 223621K->25340K(1022400K), 0.0096630
secs] [Times: user=0.03 sys=0.00, real=0.01 secs]
WARN 06:57:45 Read 2 live and 999998 tombstoned cells in
MyKeyspace.TombstoneTest (see tombstone_warn_threshold). 1000000 columns was
requested, slices=[-], delInfo={deletedAt=-9223372036854775808,
localDeletion=2147483647}
===== FULL GC AFTER =====
2014-10-07T15:57:53.746+0900: [Full GC2014-10-07T15:57:53.746+0900: [CMS:
13829K->13104K(786432K), 0.0893910 secs] 54777K->13104K(1022400K), [CMS Perm :
18786K->18785K(29924K)], 0.0895390 secs] [Times: user=0.09 sys=0.00, real=0.09
secs]
> A new config option to ignore column tombstones for RR or not
> -------------------------------------------------------------
>
> Key: CASSANDRA-8038
> URL: https://issues.apache.org/jira/browse/CASSANDRA-8038
> Project: Cassandra
> Issue Type: New Feature
> Components: Core
> Reporter: Takenori Sato
> Fix For: 2.1.1
>
> Attachments: CASSANDRA-8038-v2.txt, CASSANDRA-8038-v3.txt,
> CASSANDRA-8038.txt
>
>
> CASSANDRA-6117 addressed the death of Cassandra by column tombstones, and
> whose fix was to raise an error when reading more tombstones than a
> threshold. I think it is an emergency action, rather than a fix.
> We have had this issue for long. So I wondered, in the first place, if it is
> really necessary to collect non-gc-able tombstones, which could cause
> concurrent mode failures, and OOM eventually?
> Actually, I was surprised by the fact that Cassandra takes them into
> consideration. Rather, I prefer to raise a threshold, and tell Cassandra to
> ignore tombstones for digest calculation of RR because a repair is running
> regularly.
> I guess there are some people like me, but not all. So what about adding a
> new configuration option if Cassandra ignores column tombstones for RR or not?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
