[ 
https://issues.apache.org/jira/browse/CASSANDRA-4694?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13461689#comment-13461689
 ] 

Alexey Zotov commented on CASSANDRA-4694:
-----------------------------------------

>From description of populate_io_cache_on_flush option:
{code}
# The following setting populates the page cache on memtable flush and 
compaction
# WARNING: Enable this setting only when the whole node's data fits in memory.
# Defaults to: false
# populate_io_cache_on_flush: false
{code}

I'll try to describe my thoughts. Lets imagine that you have a few column 
families: cf1 and cf2. 
cf1 has a huge amount of data that doesn't fit in memory. Most of queries to 
cf1 use only small part of data (so-called hot data). You can use row cache for 
processing all these queries. 
cf2 has a acceptable size of data for in-memory storing. You can use 
populate_io_cache_on_flush option for fast reading of sstables from IO cache 
(from memory in fact). 

You are right about fix versions. I can migrate it to 1.2 version if it's 
needed. Should I do that?
                
> populate_io_cache_on_flush option should be configurable for each column 
> family independently
> ---------------------------------------------------------------------------------------------
>
>                 Key: CASSANDRA-4694
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-4694
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Core
>            Reporter: Alexey Zotov
>            Assignee: Alexey Zotov
>            Priority: Minor
>              Labels: cli, core, documentation
>             Fix For: 1.1.6
>
>         Attachments: cassandra-1.1-4694-populate_io_cache.txt
>
>
> I suggest to configure populate_io_cache_on_flush option for each column 
> family. It should be configurable from cassandra-cli and should be stored in 
> System keyspace. 
> That could be useful if you have a few column families inside single keyspace 
> and you need to fit in memory only one of them.
> Patch has been attached. I've been testing it on pseudo-cluster using ccm. So 
> I don't have fully confidence about lack of bugs. Please carefully review 
> that code.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to