[
https://issues.apache.org/jira/browse/HADOOP-11044?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14118486#comment-14118486
]
Gary Steelman commented on HADOOP-11044:
----------------------------------------
[~sandyr] that's unfortunate. There's a couple ways I see to handle this then:
1) Add new methods to return long types and also store longs internally,
increment them when the ints are incremented. This requires more memory per
Statistic object.
2) Add new methods to return long types and switch the internal types to long
and simply cast the internal longs to ints when returned via the int methods.
Mark the int methods deprecated, then remove them after next release. This puts
overflow problem onto the int type methods.
If we go option 2, I think the API ends up cleaner in the end, but we do
introduce a non-backwards compatible change. It might be that option 1 is the
best in this case.
> FileSystem counters can overflow for large number of readOps, largeReadOps,
> writeOps
> ------------------------------------------------------------------------------------
>
> Key: HADOOP-11044
> URL: https://issues.apache.org/jira/browse/HADOOP-11044
> Project: Hadoop Common
> Issue Type: Improvement
> Affects Versions: 2.5.0, 2.4.1
> Reporter: Swapnil Daingade
> Priority: Minor
>
> The org.apache.hadoop.fs.FileSystem.Statistics.StatisticsData class defines
> readOps, largeReadOps, writeOps as int. Also the The
> org.apache.hadoop.fs.FileSystem.Statistics class has methods like
> getReadOps(), getLargeReadOps() and getWriteOps() that return int. These int
> values can overflow if the exceed 2^31-1 showing negative values. It would be
> nice if these can be changed to long.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
