cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Benedict (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-7549) Heavy Disk Read I/O
Date Tue, 15 Jul 2014 12:11:04 GMT


Benedict commented on CASSANDRA-7549:

Can you clarify what this ticket is hoping to achieve? "Too small" compared to what? The default
Cassandra block size is 64Kb, and since currently reading data in a block requires a sequential
scan of data preceding the target record, reading ~40Kb is about correct for reading a very
small row with an 8Kb read ahead, and zero page cache utilisation (i.e. completely random
reads for a substantially larger than memory dataset). Increasing block size is likely to
reduce performance for this kind of workload, not improve it.

> Heavy Disk Read I/O
> -------------------
>                 Key: CASSANDRA-7549
>                 URL:
>             Project: Cassandra
>          Issue Type: Improvement
>         Environment: Cassandra 2.0.6
>            Reporter: Hanson
> We observed heavy disk Read I/O, sometimes almost ~100% disk I/O %util. The block size
for Disk Read seems too small per “iostat”: 
> 	- DB Query: ~40KB per read
> 	- SSTables Compaction : ~120KB per read
> Could it use larger block size for Disk Read? (from Cassandra or OS disk driver tuning)

This message was sent by Atlassian JIRA

View raw message