You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ambari.apache.org by "Andrew Onischuk (JIRA)" <ji...@apache.org> on 2014/01/16 15:05:22 UTC
[jira] [Updated] (AMBARI-4315) Proposed changes to new resource
mgmt logging
[ https://issues.apache.org/jira/browse/AMBARI-4315?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Andrew Onischuk updated AMBARI-4315:
------------------------------------
Attachment: AMBARI-4315.patch
> Proposed changes to new resource mgmt logging
> ---------------------------------------------
>
> Key: AMBARI-4315
> URL: https://issues.apache.org/jira/browse/AMBARI-4315
> Project: Ambari
> Issue Type: Bug
> Reporter: Andrew Onischuk
> Assignee: Andrew Onischuk
> Fix For: 1.5.0
>
> Attachments: AMBARI-4315.patch
>
>
> Here is a snippet of current logging.
> 1) Is the "resource_management" needed? Maybe for DEBUG, but for INFO as well?
> {code}
> 2014-01-10 00:51:02,168 - resource_management - DEBUG - Running resource Directory['/var/log/hadoop/hdfs']
> 2014-01-10 00:51:02,168 - resource_management - DEBUG - Performing action create on Directory['/var/log/hadoop/hdfs']
> 2014-01-10 00:51:02,168 - resource_management.resource - DEBUG - New resource Execute['true']: {}
> 2014-01-10 00:51:02,168 - resource_management - DEBUG - Running resource Execute['true']
> 2014-01-10 00:51:02,168 - resource_management - DEBUG - Performing action run on Execute['true']
> 2014-01-10 00:51:02,168 - resource_management.provider - INFO - Executing Execute['true']
> 2014-01-10 00:51:02,179 - resource_management.resource - DEBUG - New resource Execute['export HADOOP_LIBEXEC_DIR=/usr/lib/hadoop/libexec && /usr/lib/hadoop/sbin/hadoop-daemon.sh --config /etc/hadoop/conf start datanode']: {'not_if': u'ls /var/run/hadoop/hdfs/hadoop-hdfs-datanode.pid >/dev/null 2>&1 && ps `cat /var/run/hadoop/hdfs/hadoop-hdfs-datanode.pid` >/dev/null 2>&1', 'user': u'hdfs'}
> 2014-01-10 00:51:02,179 - resource_management - DEBUG - Running resource Execute['export HADOOP_LIBEXEC_DIR=/usr/lib/hadoop/libexec && /usr/lib/hadoop/sbin/hadoop-daemon.sh --config /etc/hadoop/conf start datanode']
> 2014-01-10 00:51:02,191 - resource_management - DEBUG - Performing action run on Execute['export HADOOP_LIBEXEC_DIR=/usr/lib/hadoop/libexec && /usr/lib/hadoop/sbin/hadoop-daemon.sh --config /etc/hadoop/conf start datanode']
> 2014-01-10 00:51:02,191 - resource_management.provider - INFO - Executing Execute['export HADOOP_LIBEXEC_DIR=/usr/lib/hadoop/libexec && /usr/lib/hadoop/sbin/hadoop-daemon.sh --config /etc/hadoop/conf start datanode']
> {code}
> 2) If not, maybe a reorg like this to remove some whitespace and tighten-up the layout.
> {code}
> 2014-01-06 23:48:18,627 DEBUG [resource_management] Performing action run on Execute['/etc/init.d/httpd stop']
> 2014-01-06 23:48:18,627 INFO Executing Execute['/etc/init.d/httpd stop']
> 2014-01-06 23:48:18,647 DEBUG [resource_management] New resource Execute['grep -E 'KeepAlive (On|Off)' /etc/httpd/conf/httpd.conf && sed -i 's/KeepAlive Off/KeepAlive On/' /etc/httpd/conf/httpd.conf || echo 'KeepAlive On' >> /etc/httpd/conf/httpd.conf']: {}
> 2014-01-06 23:48:18,647 DEBUG [resource_management] Running resource Execute['grep -E 'KeepAlive (On|Off)' /etc/httpd/conf/httpd.conf && sed -i 's/KeepAlive Off/KeepAlive On/' /etc/httpd/conf/httpd.conf || echo 'KeepAlive On' >> /etc/httpd/conf/httpd.conf']
> 2014-01-06 23:48:18,648 DEBUG [resource_management] Performing action run on Execute['grep -E 'KeepAlive (On|Off)' /etc/httpd/conf/httpd.conf && sed -i 's/KeepAlive Off/KeepAlive On/' /etc/httpd/conf/httpd.conf || echo 'KeepAlive On' >> /etc/httpd/conf/httpd.conf']
> 2014-01-06 23:48:18,648 INFO Executing Execute['grep -E 'KeepAlive (On|Off)' /etc/httpd/conf/httpd.conf && sed -i 's/KeepAlive Off/KeepAlive On/' /etc/httpd/conf/httpd.conf || echo 'KeepAlive On' >> /etc/httpd/conf/httpd.conf']
> 2014-01-06 23:48:18,663 DEBUG [resource_management] New resource Execute['/etc/init.d/httpd start']: {}
> 2014-01-06 23:48:18,663 DEBUG [resource_management] Running resource Execute['/etc/init.d/httpd start']
> 2014-01-06 23:48:18,663 DEBUG [resource_management] Performing action run on Execute['/etc/init.d/httpd start']
> 2014-01-06 23:48:18,663 INFO Executing Execute['/etc/init.d/httpd start']
> {code}
> 3) On the UI side, it would be great to turn off seeing debug. It's useful to be there in case you need those details, but optionally show/don't show in UI. Because then you would end up with this which is very manageable for a user to read (this is the above logging w/ DEBUG removed).
> {code}
> 2014-01-06 23:48:18,627 INFO Executing Execute['/etc/init.d/httpd stop']
> 2014-01-06 23:48:18,648 INFO Executing Execute['grep -E 'KeepAlive (On|Off)' /etc/httpd/conf/httpd.conf && sed -i 's/KeepAlive Off/KeepAlive On/' /etc/httpd/conf/httpd.conf || echo 'KeepAlive On' >> /etc/httpd/conf/httpd.conf']
> 2014-01-06 23:48:18,663 INFO Executing Execute['/etc/init.d/httpd start']
> {code}
> 4) Also I propose when you install a package, you also give the detail of the command used. That's useful to a user. Right now, we show this:
> {code}
> 2014-01-10 00:50:48,480 - resource_management.provider - INFO Installing package ambari-log4j
> {code}
> Would be good to give the actual + specific yum/zypper/etc command being called, even in INFO:
> {code}
> 2014-01-10 00:50:48,480 INFO Installing package ambari-log4j [‘/usr/bin/yum -d 0 -e 0 -y install ambari-log4j’]
> {code}
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)