[
https://issues.apache.org/jira/browse/HBASE-18248?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16058697#comment-16058697
]
Allan Yang commented on HBASE-18248:
------------------------------------
Just a minor suggest, why do we need to pass a conf every time?
{code}
/**
* Get singleton instance.
* TODO this would be better off scoped to a single daemon
*/
- public static synchronized TaskMonitor get() {
+ public static synchronized TaskMonitor get(Configuration conf) {
if (instance == null) {
- instance = new TaskMonitor();
+ instance = new TaskMonitor(conf);
}
return instance;
}
{code}
TaskMonitor is a singleton, can we consider create a configuration when
creating TaskMonitor, so we don't need to pass a conf every time.
{code}
TaskMonitor() {
Configuration conf = HBaseConfiguration.create();
......
}
{code}
> Warn if monitored task has been tied up beyond a configurable threshold
> -----------------------------------------------------------------------
>
> Key: HBASE-18248
> URL: https://issues.apache.org/jira/browse/HBASE-18248
> Project: HBase
> Issue Type: Improvement
> Reporter: Andrew Purtell
> Assignee: Andrew Purtell
> Fix For: 2.0.0, 3.0.0, 1.4.0, 1.3.2
>
> Attachments: HBASE-18248-branch-1.3.patch,
> HBASE-18248-branch-1.patch, HBASE-18248-branch-2.patch, HBASE-18248.patch
>
>
> Warn if monitored task has been tied up beyond a configurable threshold. We
> especially want to do this for RPC tasks. Use a separate threshold for
> warning about stuck RPC tasks versus other types of tasks.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
