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-1470) Rework FSInputChecker and FSOutputSummer to support checksum code sharing between ChecksumFileSystem and block level crc dfs
Date Thu, 28 Jun 2007 18:36:04 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-1470?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12508896

Doug Cutting commented on HADOOP-1470:

Sameer, you suggested yesterday that we might use a larger, multi-chunk buffer in the InputChecker
to address your concerns about excessive seeks when multiple threads use pread on a shared
file handle with a small chunk size.  I agree that we need a larger buffer somewhere.  Earlier
I've suggested that implementations should use a larger buffer internally.  But we could,
as you suggest, instead make InputChecker's buffer larger, with a policy that data is checksummed
as late as possible.  Each chunk within the buffer would be checksummed immediately prior
to returning any data from it.  Implementing that could be a bit hairy and hence bug prone.
 It might be easiest to implement this as two buffers: one, multi-chunk and un-checksummed,
and one checksummed single-chunk, the latter filled from the former.  As an optimization,
they could even share buffer memory, so that refilling the checksummed buffer would involve
just checksumming some data and resetting its start position.  But, in the code, this would
be abstracted as two cascaded buffers, to address complexity.  Does that make sense?

> Rework FSInputChecker and FSOutputSummer to support checksum code sharing between ChecksumFileSystem
and block level crc dfs
> ----------------------------------------------------------------------------------------------------------------------------
>                 Key: HADOOP-1470
>                 URL: https://issues.apache.org/jira/browse/HADOOP-1470
>             Project: Hadoop
>          Issue Type: Improvement
>          Components: fs
>    Affects Versions: 0.12.3
>            Reporter: Hairong Kuang
>            Assignee: Hairong Kuang
>             Fix For: 0.14.0
>         Attachments: GenericChecksum.patch, genericChecksum.patch, GenericChecksum1.patch,
> Comment from Doug in HADOOP-1134:
> I'd prefer it if the CRC code could be shared with CheckSumFileSystem. In particular,
it seems to me that FSInputChecker and FSOutputSummer could be extended to support pluggable
sources and sinks for checksums, respectively, and DFSDataInputStream and DFSDataOutputStream
could use these. Advantages of this are: (a) single implementation of checksum logic to debug
and maintain; (b) keeps checksumming as close to possible to data generation and use. This
patch computes checksums after data has been buffered, and validates them before it is buffered.
We sometimes use large buffers and would like to guard against in-memory errors. The current
checksum code catches a lot of such errors. So we should compute checksums after minimal buffering
(just bytesPerChecksum, ideally) and validate them at the last possible moment (e.g., through
the use of a small final buffer with a larger buffer behind it). I do not think this will
significantly affect performance, and data integrity is a high priority. 

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

View raw message