[
https://issues.apache.org/jira/browse/HDFS-4817?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13659670#comment-13659670
]
Todd Lipcon commented on HDFS-4817:
-----------------------------------
Some comments on the API:
Setting it in the client configuration is nice to be able to set a default (eg
experiment with a setting on an MR job without changing code), but it's not
useful if you want to set different settings on each stream, which is quite
likely in an application like HBase. The setting in DistributedFileSystem is
also not really great since we have the FileSystem cache -- in any
multi-threaded application, you're likely to share DFS instances across
threads. So, your attempt to set the caching strategy on the DFS level will not
be very reliable either.
I'd like to see this API either as additional calls that can be made on the
stream (like how fadvise() works in the posix API), or as extra arguments to
create/open calls.
> make HDFS advisory caching configurable on a per-file basis
> -----------------------------------------------------------
>
> Key: HDFS-4817
> URL: https://issues.apache.org/jira/browse/HDFS-4817
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: hdfs-client
> Affects Versions: 3.0.0
> Reporter: Colin Patrick McCabe
> Assignee: Colin Patrick McCabe
> Priority: Minor
> Attachments: HDFS-4817.001.patch
>
>
> HADOOP-7753 and related JIRAs introduced some performance optimizations for
> the DataNode. One of them was readahead. When readahead is enabled, the
> DataNode starts reading the next bytes it thinks it will need in the block
> file, before the client requests them. This helps hide the latency of
> rotational media and send larger reads down to the device. Another
> optimization was "drop-behind." Using this optimization, we could remove
> files from the Linux page cache after they were no longer needed.
> Using {{dfs.datanode.drop.cache.behind.writes}} and
> {{dfs.datanode.drop.cache.behind.reads}} can improve performance
> substantially on many MapReduce jobs. In our internal benchmarks, we have
> seen speedups of 40% on certain workloads. The reason is because if we know
> the block data will not be read again any time soon, keeping it out of memory
> allows more memory to be used by the other processes on the system. See
> HADOOP-7714 for more benchmarks.
> We would like to turn on these configurations on a per-file or per-client
> basis, rather than on the DataNode as a whole. This will allow more users to
> actually make use of them. It would also be good to add unit tests for the
> drop-cache code path, to ensure that it is functioning as we expect.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
