[
https://issues.apache.org/jira/browse/HDFS-10475?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15491551#comment-15491551
]
Zhe Zhang commented on HDFS-10475:
----------------------------------
Nice discussion guys.
A quick note from looking at the code:
{code}
public Writable call(RPC.Server server, String protocol,
Writable writableRequest, long receiveTime) throws Exception {
...
long startTime = Time.now();
int qTime = (int) (startTime - receiveTime);
...
try {
server.rpcDetailedMetrics.init(protocolImpl.protocolClass);
result = service.callBlockingMethod(methodDescriptor, null, param);
...
} finally {
int processingTime = (int) (Time.now() - startTime);
...
server.updateMetrics(detailedMetricsName, qTime, processingTime);
{code}
So the reported queue time is from when the RPC request enters *RPC queue* to
the time it exits the queue. The processing time is from the time it exists RPC
queue until the request is completed. The proposed _lock time_ will be part of
the _processing time_.
> Adding metrics for long FSNamesystem read and write locks
> ---------------------------------------------------------
>
> Key: HDFS-10475
> URL: https://issues.apache.org/jira/browse/HDFS-10475
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: namenode
> Reporter: Xiaoyu Yao
> Assignee: Erik Krogen
>
> This is a follow up of the comment on HADOOP-12916 and
> [here|https://issues.apache.org/jira/browse/HDFS-9924?focusedCommentId=15310837&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-15310837]
> add more metrics and WARN/DEBUG logs for long FSD/FSN locking operations on
> namenode similar to what we have for slow write/network WARN/metrics on
> datanode.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
