[
https://issues.apache.org/jira/browse/CASSANDRA-6370?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13825405#comment-13825405
]
Alain RODRIGUEZ commented on CASSANDRA-6370:
--------------------------------------------
I didn't try to list any table here.
I heard was thrift / cql were abstractions of the same data and that we could
continuing to use both. I thought cassandra-cli was also compatible with cql
tables. Even more, I had no error or warning while running my cassandra-cli
command.
In my point of view (which was wrong at this time, I agree now) this was a
"normal" usage of cassandra that resulted into a bug in my productions servers.
My error is now fixed as you can read at the end of the description. My point
here is to help the Cassandra team to make Cassandra more robust to avoid more
issues of this kind.
Do whatever you want with this report, but there is no need of bashing me, this
wasn't trivial, and the fact you hide the cql-created tables doesn't help in
any way since I didn't list them.
This happen to me for being an early Cassandra adopter who use to change the
schema using cassandra-cli.
> Updating cql created table through cassandra-cli transform it into a compact
> storage table
> ------------------------------------------------------------------------------------------
>
> Key: CASSANDRA-6370
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6370
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Alain RODRIGUEZ
> Assignee: Sylvain Lebresne
> Priority: Critical
>
> To reproduce :
> echo "CREATE TABLE test (aid int, period text, event text, viewer text,
> PRIMARY KEY (aid, period, event, viewer) );" | cqlsh -kmykeyspace;
> echo "describe table test;" | cqlsh -kmykeyspace;
> Output >
> CREATE TABLE test (
> aid int,
> period text,
> event text,
> viewer text,
> PRIMARY KEY (aid, period, event, viewer)
> ) WITH
> bloom_filter_fp_chance=0.010000 AND
> caching='KEYS_ONLY' AND
> comment='' AND
> dclocal_read_repair_chance=0.000000 AND
> gc_grace_seconds=864000 AND
> read_repair_chance=0.100000 AND
> replicate_on_write='true' AND
> populate_io_cache_on_flush='false' AND
> compaction={'class': 'SizeTieredCompactionStrategy'} AND
> compression={'sstable_compression': 'SnappyCompressor'};
> Then do :
> echo "update column family test with dclocal_read_repair_chance = 0.1;" |
> cassandra-cli -kmykeyspace
> And finally again : echo "describe table test;" | cqlsh -kmykeyspace;
> Output >
> CREATE TABLE test (
> aid int,
> column1 text,
> column2 text,
> column3 text,
> column4 text,
> value blob,
> PRIMARY KEY (aid, column1, column2, column3, column4)
> ) WITH COMPACT STORAGE AND
> bloom_filter_fp_chance=0.010000 AND
> caching='KEYS_ONLY' AND
> comment='' AND
> dclocal_read_repair_chance=0.100000 AND
> gc_grace_seconds=864000 AND
> read_repair_chance=0.100000 AND
> replicate_on_write='true' AND
> populate_io_cache_on_flush='false' AND
> compaction={'class': 'SizeTieredCompactionStrategy'} AND
> compression={'sstable_compression': 'SnappyCompressor'};
> This is quite annoying in production. If it is happening to you:
> UPDATE system.schema_columnfamilies SET column_aliases =
> '["period","event","viewer"]' WHERE keyspace_name='mykeyspace' AND
> columnfamily_name='test'; should help restoring the table. (Thanks Sylvain
> for this information.)
--
This message was sent by Atlassian JIRA
(v6.1#6144)
