[
https://issues.apache.org/jira/browse/HDFS-17189?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ConfX updated HDFS-17189:
-------------------------
Description:
The garbage collection data buffer size equals to
`dfs.namenode.gc.time.monitor.observation.window.ms` divided by
`dfs.namenode.gc.time.monitor.sleep.interval.ms` plus 2. When the observation
window is a large value, the caculation overflows, resulting in a negative
buffer size.
To reproduce:
1. set `dfs.namenode.gc.time.monitor.observation.window.ms` to 945099495m
2. run `mvn surefire:test
-Dtest=org.apache.hadoop.hdfs.TestHDFSFileSystemContract#testWriteReadAndDeleteHalfABlock`
We create a PR that provides a fix by checking the computed buffer size is not
only within `128 * 1024` but also greater than 0.
was:
The garbage collection data buffer size equals to
`dfs.namenode.gc.time.monitor.observation.window.ms` divided by
`dfs.namenode.gc.time.monitor.sleep.interval.ms` plus 2. When the observation
window is a large value, the caculation overflows, resulting in a negative
buffer size.
To reproduce:
1. set `dfs.namenode.gc.time.monitor.observation.window.ms` to 945099495m
2. run `mvn surefire:test
-Dtest=org.apache.hadoop.hdfs.TestHDFSFileSystemContract#testWriteReadAndDeleteHalfABlock`
This PR provides a fix by checking the computed buffer size is not only within
`128 * 1024` but also greater than 0.
> GcTimeMonitor crashes with NegativeArraySizeException during initialization
> ---------------------------------------------------------------------------
>
> Key: HDFS-17189
> URL: https://issues.apache.org/jira/browse/HDFS-17189
> Project: Hadoop HDFS
> Issue Type: Bug
> Affects Versions: 3.3.6
> Reporter: ConfX
> Priority: Major
> Labels: pull-request-available
>
> The garbage collection data buffer size equals to
> `dfs.namenode.gc.time.monitor.observation.window.ms` divided by
> `dfs.namenode.gc.time.monitor.sleep.interval.ms` plus 2. When the observation
> window is a large value, the caculation overflows, resulting in a negative
> buffer size.
> To reproduce:
> 1. set `dfs.namenode.gc.time.monitor.observation.window.ms` to 945099495m
> 2. run `mvn surefire:test
> -Dtest=org.apache.hadoop.hdfs.TestHDFSFileSystemContract#testWriteReadAndDeleteHalfABlock`
> We create a PR that provides a fix by checking the computed buffer size is
> not only within `128 * 1024` but also greater than 0.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
