hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sameer Paranjpye (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-1134) Block level CRCs in HDFS
Date Thu, 29 Mar 2007 19:15:25 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-1134?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12485308

Sameer Paranjpye commented on HADOOP-1134:

> The same way we do today: we don't. 

When the HDFS client encounters a checksum error, it doesn't know whether it is the data or
the checksum that is corrupt. But it takes care to switch both the data block and the checksum
block that it is reading. So while we don't *know* whether the checksum is corrupt, we certainly
don't assume that it's not.

Computing CRCs is cheap, every job that we run computes them on input and on output, and it
hasn't been a burden so far. We can benchmark how long it takes to compute checksums for a
block, but I doubt it would add significantly to the time for upgrade.

> This patch should confine itself to removing checksums from the HDFS namespace 

Perhaps, but even so, the removal should be done correctly. The system as it exists today
deals with the possibility of checksum corruption, it should do so during and after the upgrade.
Not doing so would be a regression.

> Block level CRCs in HDFS
> ------------------------
>                 Key: HADOOP-1134
>                 URL: https://issues.apache.org/jira/browse/HADOOP-1134
>             Project: Hadoop
>          Issue Type: New Feature
>          Components: dfs
>            Reporter: Raghu Angadi
>         Assigned To: Raghu Angadi
> Currently CRCs are handled at FileSystem level and are transparent to core HDFS. See
recent improvement HADOOP-928 ( that can add checksums to a given filesystem ) regd more about
it. Though this served us well there a few disadvantages :
> 1) This doubles namespace in HDFS ( or other filesystem implementations ). In many cases,
it nearly doubles the number of blocks. Taking namenode out of CRCs would nearly double namespace
performance both in terms of CPU and memory.
> 2) Since CRCs are transparent to HDFS, it can not actively detect corrupted blocks. With
block level CRCs, Datanode can periodically verify the checksums and report corruptions to
namnode such that name replicas can be created.
> We propose to have CRCs maintained for all HDFS data in much the same way as in GFS.
I will update the jira with detailed requirements and design. This will include same guarantees
provided by current implementation and will include a upgrade of current data.

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

View raw message