cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yanis Biziuk (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-4233) overlapping sstables in leveled compaction strategy
Date Thu, 07 Jun 2012 09:37:23 GMT


Yanis Biziuk commented on CASSANDRA-4233:

Stack trace of this error

ERROR 04:09:21,216 Exception in thread Thread[CompactionExecutor:410,1,main]
java.lang.AssertionError: Last written key DecoratedKey(144093708553026671072703572185472027831,
00000000003eef36) >= current key DecoratedKey(50602742485951681279829043048319220646, 00000000004506bf)
writing into /cassandra/data/Server/Messages/Server-Messages-tmp-hd-5380-Data.db
	at org.apache.cassandra.db.compaction.CompactionTask.execute(
	at org.apache.cassandra.db.compaction.LeveledCompactionTask.execute(
	at org.apache.cassandra.db.compaction.CompactionManager$1.runMayThrow(
	at java.util.concurrent.Executors$ Source)
	at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source)
	at Source)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source)
	at java.util.concurrent.ThreadPoolExecutor$ Source)
	at Source)

Linux version 2.6.32-5-amd64 (Debian 2.6.32-45)
32GB ram, 24 threads (12 cores)

We are upgraded Cassandra server from repository(config file was overwritten, but we change
it later). 
And we REcreated keyspace with all CF (we in not a production stage yet).

all us CFs
UPDATE COLUMN FAMILY Messages WITH compression_options={sstable_compression:SnappyCompressor,
UPDATE COLUMN FAMILY Messages WITH compaction_strategy=LeveledCompactionStrategy AND compaction_strategy_options={sstable_size_in_mb:
UPDATE COLUMN FAMILY Messages WITH gc_grace = 0;

This CF has many inserts in one time. 
Over 1000-3000 keys with 9 columns each in one sec.
And this CF has 4 secondary indexes (numbers ~ from 1 to 20)

More info about Cassandra server
partitioner: org.apache.cassandra.dht.RandomPartitioner
key_cache_size_in_mb - default
row_cache_size_in_mb - default
row_cache_provider: SerializingCacheProvider


Thanks for any help.

> overlapping sstables in leveled compaction strategy
> ---------------------------------------------------
>                 Key: CASSANDRA-4233
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>    Affects Versions: 1.1.0
>            Reporter: Jonathan Ellis
>            Assignee: Sylvain Lebresne
>             Fix For: 1.1.1
>         Attachments: 0001-Make-sure-Leveled-compaction-always-use-the-right-min-.txt,
4233-assert.txt, 4233-v2.txt, 4233-v3.txt, 4233-v4.txt, system.log.bz2
> CASSANDRA-4142 introduces test failures, that are caused by overlapping tables within
a level, which Shouldn't Happen.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


View raw message