[
https://issues.apache.org/jira/browse/HDFS-8163?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Arpit Agarwal updated HDFS-8163:
--------------------------------
Resolution: Fixed
Fix Version/s: 2.7.1
Target Version/s: (was: 2.7.1)
Hadoop Flags: Reviewed
Status: Resolved (was: Patch Available)
Thanks for the review Jing.
Fixed the formatting and committed to trunk, branch-2 and branch-2.7. Here is
the delta:
{code:java}
- // assigned/read by the actor thread. Thus they should be declared as vol
- // to make sure the "happens-before" consistency.
- @VisibleForTesting volatile long nextBlockReportTime = monotonicNow();
- @VisibleForTesting volatile long nextHeartbeatTime = monotonicNow();
- @VisibleForTesting boolean resetBlockReportTime = true;
+ // assigned/read by the actor thread.
+ @VisibleForTesting
+ volatile long nextBlockReportTime = monotonicNow();
+
+ @VisibleForTesting
+ volatile long nextHeartbeatTime = monotonicNow();
+
+ @VisibleForTesting
+ boolean resetBlockReportTime = true;
{code}
> Using monotonicNow for block report scheduling causes test failures on
> recently restarted systems
> -------------------------------------------------------------------------------------------------
>
> Key: HDFS-8163
> URL: https://issues.apache.org/jira/browse/HDFS-8163
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: datanode
> Affects Versions: 2.6.1
> Reporter: Arpit Agarwal
> Assignee: Arpit Agarwal
> Priority: Blocker
> Fix For: 2.7.1
>
> Attachments: HDFS-8163.01.patch, HDFS-8163.02.patch,
> HDFS-8163.03.patch
>
>
> {{BPServiceActor#blockReport}} has the following check:
> {code}
> List<DatanodeCommand> blockReport() throws IOException {
> // send block report if timer has expired.
> final long startTime = monotonicNow();
> if (startTime - lastBlockReport <= dnConf.blockReportInterval) {
> return null;
> }
> {code}
> Many tests trigger an immediate block report via
> {{BPServiceActor#triggerBlockReportForTests}} which sets {{lastBlockReport =
> 0}}. However if the machine was restarted recently then startTime may be less
> than {{dnConf.blockReportInterval}} and the block report is not sent.
> {{Time#monotonicNow}} uses {{System#nanoTime}} which represents time elapsed
> since an arbitrary origin. The time should be used only for comparison with
> other values returned by {{System#nanoTime}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
