Mailing-List: contact commits-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cassandra.apache.org
Message-ID: <24866523.124181275456158411.JavaMail.jira@thor>
Date: Wed, 2 Jun 2010 01:22:38 -0400 (EDT)
From: "Jon Hermes (JIRA)" <jira@apache.org>
To: commits@cassandra.apache.org
Subject: [jira] Issue Comment Edited: (CASSANDRA-1129) Using
 KeysCached="xx%" results in Key cache capacity: 1
In-Reply-To: <4176155.38861274808204211.JavaMail.jira@thor>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/CASSANDRA-1129?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12874427#action_12874427 ] 

Jon Hermes edited comment on CASSANDRA-1129 at 6/2/10 1:21 AM:
---------------------------------------------------------------

(I didn't catch the failures because I had run CompactionsTest instead of CompactionsPurgeTest by accident.)

This test wasn't intended to guarantee the math in estimating keys works, only to show that flushing and compaction affect the keyCacheSize dynamically.
The number of keys in the indexSummary has been changed in trunk in some way that I'm not seeing immediately.
This patch extends its range of sane replies to be either 2X or X -- that both keys are added to the indexsummary (0.7) or that only one is (0.6).

      was (Author: jhermes):
    (I didn't catch the failures because I had run CompactionsTest instead of CompactionsPurgeTest by accident.)

This test wasn't intended to guarantee the math in estimating keys works, only to show that flushing and compaction affect the keyCacheSize dynamically.
The number of estimated keys coming back has been changed in trunk in some way that I'm not seeing immediately.
This patch extends its range of sane replies to be either 2X or X -- that both keys are added to the indexsummary (0.7) or that only one is (0.6).
  
> Using KeysCached="xx%" results in Key cache capacity: 1
> -------------------------------------------------------
>
>                 Key: CASSANDRA-1129
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-1129
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.6.1
>            Reporter: Ran Tavory
>            Priority: Minor
>             Fix For: 0.6.3
>
>         Attachments: 0001-CASSANDRA1129.patch, 0002-CASSANDRA1129.patch, 0003-CASSANDRA1129.patch, KvAds-84.zip, TOTRUNK-2-CASSANDRA1129.patch, TOTRUNK-CASSANDRA1129.patch
>
>
> I don't know if this is a general bug or only something related to my instance, but for me (v0.6.1) I've noticed that when defining KeysCached="50%" (or KeysCached="100%" and I didn't test other values with %) then cfstats reports Key cache capacity: 1
>       <ColumnFamily CompareWith="BytesType" Name="KvAds"
>         KeysCached="100%"
>         RowsCached="10000"
>         />
>                 Column Family: KvAds
>                 SSTable count: 7
>                 Space used (live): 797535964
>                 Space used (total): 797535964
>                 Memtable Columns Count: 42292
>                 Memtable Data Size: 10514176
>                 Memtable Switch Count: 24
>                 Read Count: 2563704
>                 Read Latency: 4.590 ms.
>                 Write Count: 1963804
>                 Write Latency: 0.025 ms.
>                 Pending Tasks: 0
>                 Key cache capacity: 1
>                 Key cache size: 1
>                 Key cache hit rate: 0.0
>                 Row cache capacity: 10000
>                 Row cache size: 10000
>                 Row cache hit rate: 0.2206178354382234
>                 Compacted row minimum size: 386
>                 Compacted row maximum size: 9808
>                 Compacted row mean size: 616
> I'll attach one of the sstable files from this CF

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