Mailing-List: contact hdfs-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Date: Mon, 5 Jun 2017 02:37:04 +0000 (UTC)
From: "Yiqun Lin (JIRA)" <jira@apache.org>
To: hdfs-issues@hadoop.apache.org
Message-ID: <JIRA.13076940.1496428058000.362229.1496630224070@Atlassian.JIRA>
In-Reply-To: <JIRA.13076940.1496428058000@Atlassian.JIRA>
References: <JIRA.13076940.1496428058000@Atlassian.JIRA> <JIRA.13076940.1496428058203@jira-lw-us.apache.org>
Subject: [jira] [Comment Edited] (HDFS-11919) Ozone: SCM: TestNodeManager
 takes too long to execute
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Mon, 05 Jun 2017 02:37:08 -0000


    [ https://issues.apache.org/jira/browse/HDFS-11919?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16036476#comment-16036476 ] 

Yiqun Lin edited comment on HDFS-11919 at 6/5/17 2:36 AM:
----------------------------------------------------------

Attach the patch. Did the following improvements in test:

* Set {{OZONE_SCM_HEARTBEAT_PROCESS_INTERVAL_MS}} to 100ms since we are just for testing, no need to use default value(3s).
* Combine the test methods {{testScmDetectStale/DeadNode}} into one single test {{testScmDetectStaleAndDeadNode}}. Because the logic of {{testScmDetectStale/DeadNode}} is almost same, we can just combine this.
* Improve the logic in {{testScmLogsHeartbeatFlooding}}, we can just check the output rather than sleep 6s first. If not contains, it will retry again after 500ms interval time.

Tested in my local, it totally costs around 1 min.


was (Author: linyiqun):
Attach the patch. Did the following improvements in test:

* Set {{OZONE_SCM_HEARTBEAT_PROCESS_INTERVAL_MS}} to 100ms since we are just for testing, no need to use default value(3s).
* Combine the test methods {{testScmDetectStale/DeadNode}} into one single test {{testScmDetectStaleAndDeadNode}}. Because the logic of {{testScmDetectStale/DeadNode}} is almost same, we can just combine this.
* Improve the logic in {{testScmLogsHeartbeatFlooding}}, we can just check the output rather than sleep 6s first. If not contains, it will retry again after 500ms interval time.

> Ozone: SCM: TestNodeManager takes too long to execute
> -----------------------------------------------------
>
>                 Key: HDFS-11919
>                 URL: https://issues.apache.org/jira/browse/HDFS-11919
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: ozone
>    Affects Versions: HDFS-7240
>            Reporter: Anu Engineer
>            Assignee: Yiqun Lin
>            Priority: Trivial
>         Attachments: HDFS-11919-HDFS-7240.001.patch
>
>
> On my laptop it takes 97.645 seconds to execute this test. We should explore if we can make this test run faster. 


--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org