cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Stu Hood (JIRA)" <>
Subject [jira] Commented: (CASSANDRA-792) Configurable 'merge factor' for Column Families
Date Mon, 15 Feb 2010 01:21:27 GMT


Stu Hood commented on CASSANDRA-792:

> how is this different from existing min/max compaction files? sounds like just a rename
to me from that quote. 
There is one other difference: mergeFactor also determines the size of the buckets that sstables
are compacted in.

> My initial reaction is that this is not something we want to do.
I think we may be forced to. Without throttling writes to a level that compactions can sustain,
you can run into situations where you have many thousands of sstables on disk, which leads
to abysmal read performance. If people _want_ to be in this situation, then we can allow them
to disable compaction, but otherwise, compaction needs to be able to keep up.

> Configurable 'merge factor' for Column Families
> -----------------------------------------------
>                 Key: CASSANDRA-792
>                 URL:
>             Project: Cassandra
>          Issue Type: New Feature
>            Reporter: Stu Hood
>             Fix For: 0.6
>         Attachments: 0001-Use-logarithmic-bucket-sizes-and-always-compact-the-.patch,
0002-Cap-memory-usage-for-all-SSTables-being-compacted.patch, 0003-Hardcoded-per-cf-merge-factor.patch,
> Apache Lucene flushes and merges 'segments' in a manner that is quite similar to the
treatment of Cassandra's sstables.
> One interesting and powerful tunable that Lucene provides is 'mergeFactor':
> EDIT: better link:

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message