[
https://issues.apache.org/jira/browse/CASSANDRA-47?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12771956#action_12771956
]
Jonathan Ellis commented on CASSANDRA-47:
-----------------------------------------
A very straightforward way to compress would be on a per-[column]index-block
basis. for large rows this should work well, and requires only minor changes
to sstable reading/writing. compression across multiple rows would work better
for small rows, but requires more invasive changes, and adds a lot of
complexity for large rows (do you break the row across multiple blocks? or
decompress the entire row into memory every time? both options suck).
> SSTable compression
> -------------------
>
> Key: CASSANDRA-47
> URL: https://issues.apache.org/jira/browse/CASSANDRA-47
> Project: Cassandra
> Issue Type: New Feature
> Components: Core
> Reporter: Jonathan Ellis
> Assignee: Eric Evans
> Priority: Minor
>
> We should be able to do SSTable compression which would trade CPU for I/O
> (almost always a good trade).
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
