hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lars Hofhansl (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-10552) HFilePerformanceEvaluation.GaussianRandomReadBenchmark fails sometimes.
Date Sun, 16 Feb 2014 20:30:21 GMT

    [ https://issues.apache.org/jira/browse/HBASE-10552?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13902820#comment-13902820
] 

Lars Hofhansl commented on HBASE-10552:
---------------------------------------

Trunk has a different fix for this:
{code}
      if (scanner.seekTo(gaussianRandomRowBytes) < 0) {
        LOG.info("Not able to seekTo " + new String(gaussianRandomRowBytes));
        return;
      }
{code}
Which is a hack, IMHO. The reason we get -1 is because we generated a seek key before the
first key of the file.

> HFilePerformanceEvaluation.GaussianRandomReadBenchmark fails sometimes.
> -----------------------------------------------------------------------
>
>                 Key: HBASE-10552
>                 URL: https://issues.apache.org/jira/browse/HBASE-10552
>             Project: HBase
>          Issue Type: Bug
>            Reporter: Lars Hofhansl
>            Priority: Minor
>             Fix For: 0.96.2, 0.98.1, 0.99.0, 0.94.17
>
>         Attachments: 10552-0.94.txt
>
>
> GaussianRandomReadBenchmark generates seek keys by using a Gaussian distribution with
the mean of N/2 and a sigma of N/10 (N = number of rows used)  and using this key directly
to seek into the HFile. The HFile was seeded with keys from 0-N.
> This will fail if we ever generate a key < 0, which is rare, but by no means impossible.
We need to clamp the min and max values to 0 and N, resp.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)

Mime
View raw message