[
https://issues.apache.org/jira/browse/AMBARI-14603?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15091351#comment-15091351
]
Sumit Mohanty edited comment on AMBARI-14603 at 1/11/16 2:46 AM:
-----------------------------------------------------------------
It seems that the issue itself is intermittent. I noticed it a couple of times
but it was not clear why. I cleaned some code from the perspective of pattern
and added some logs.
was (Author: sumitmohanty):
It seems that reproduction is not easy. I noticed it a couple of times but it
was not clear why. I cleaned some code from the perspective of pattern and
added some logs.
> Components marked for auto-start seemingly starting even after STOP has been
> issued
> -----------------------------------------------------------------------------------
>
> Key: AMBARI-14603
> URL: https://issues.apache.org/jira/browse/AMBARI-14603
> Project: Ambari
> Issue Type: Bug
> Components: ambari-agent
> Affects Versions: 2.2.1
> Reporter: Sumit Mohanty
> Assignee: Sumit Mohanty
> Priority: Critical
> Fix For: 2.2.1
>
> Attachments: AMBARI-14603.patch
>
>
> Steps:
> # Deployed cluster with custom log and pid dirs via blueprints.
> # Changed log dirs for Storm, Ambari Metrics, Atlas services.
> # Restarted needed components.
> # Stopped Atlas service, changed pid dir, started Atlas. Atlas started
> successfully.
> # Repeated this for Storm.
> # Stopped Ambari Metrics.
> # Change pid dir.
> # Click save.
> Result: during saving Ambari Metrics service status was changed to Started
> for some time (see from 1:23). After some time status was returned to correct.
> High level commands:
> * Dec 29 18:07 command-242.json - STOP METRICS_COLLECTOR
> * Dec 29 18:08 auto_command-1451432816.json - AUTO_START METRICS_COLLECTOR
> (metrics_monitor_pid_dir - /grid/0/pid/metric_monitor)
> * Dec 29 18:13 command-246.json - RESTART METRICS_COLLECTOR
> (metrics_monitor_pid_dir - /grid/0/pid/metric_monitor1)
> * Dec 29 18:15 auto_command-1451432817.json
> From agent logs:
> * INFO 2015-12-29 23:58:01,583 ActionQueue.py:339 - After EXECUTION_COMMAND
> (STOP/INSTALL), current state of METRICS_COLLECTOR to INSTALLED
> * INFO 2015-12-30 00:15:38,135 ActionQueue.py:335 - After EXECUTION_COMMAND
> (START), current state of METRICS_COLLECTOR to STARTED
> * INFO 2015-12-30 01:19:57,773 ActionQueue.py:346 - After EXECUTION_COMMAND
> (RESTART), current state of METRICS_COLLECTOR to STARTED
> * INFO 2015-12-30 02:07:16,156 RecoveryManager.py:674 - Received
> EXECUTION_COMMAND (STOP/INSTALL), desired state of METRICS_COLLECTOR to
> INSTALLED
> * INFO 2015-12-30 02:07:35,960 ActionQueue.py:339 - After EXECUTION_COMMAND
> (STOP/INSTALL), current state of METRICS_COLLECTOR to INSTALLED
> * *Why???* INFO 2015-12-30 02:07:53,916 RecoveryManager.py:249 -
> METRICS_COLLECTOR needs recovery.
> * INFO 2015-12-30 02:09:24,185 ActionQueue.py:335 - After EXECUTION_COMMAND
> (START), current state of METRICS_COLLECTOR to STARTED
> * INFO 2015-12-30 02:13:48,732 RecoveryManager.py:685 - Received
> EXECUTION_COMMAND (RESTART), desired state of METRICS_COLLECTOR to STARTED
> * INFO 2015-12-30 02:14:54,754 ActionQueue.py:346 - After EXECUTION_COMMAND
> (RESTART), current state of METRICS_COLLECTOR to STARTED
> * *Why???* INFO 2015-12-30 02:14:55,475 RecoveryManager.py:249 -
> METRICS_COLLECTOR needs recovery.
> * INFO 2015-12-30 02:15:47,163 ActionQueue.py:335 - After EXECUTION_COMMAND
> (START), current state of METRICS_COLLECTOR to STARTED
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
