hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Doug Cutting (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-1134) Block level CRCs in HDFS
Date Tue, 03 Apr 2007 20:15:32 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-1134?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12486481

Doug Cutting commented on HADOOP-1134:

> we should have some conventions like 'checksum of 0000 followed by some magic 8 bytes
means checksum is incorrect' or some such thing

That assumes that the checksum is validated, doesn't it?  And we're not yet validating checksums
except in the client, so I don't see where such an encoding would be used.  Right now we only
need a way for the datanode to indicate either (1) that there is a checksum, and here it is;
or (b) there is no checksum.

> if CRC file does not exist, it will be treated just like a corrupt block

The client might handle these cases differently.  If the data does not match the checksum,
then odds are the data is invalid.  However if the data has no checksum then odds are good
that the data is valid.  So one might reasonably configure the client to permit reading of
data without checksums but to throw exceptions for data whose checksum does not match the

> 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