[
https://issues.apache.org/jira/browse/SOLR-12173?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Shawn Heisey updated SOLR-12173:
--------------------------------
Description:
Some log events that I think Solr should have when it starts up. All of these
should have unique logger names or text strings so that a verbose log can be
searched to find them.
* An event indicating when startup is complete. The time to log this is when
all cores are loaded/started (or inventoried for lotsOfCores functionality),
cloud information is examined and cached as required, background threads
started, etc.
** The log message should included JVM uptime and elapsed time since Solr (the
webapp) started.
** If somebody were to hot-deploy Solr into a running container, the JVM
uptime would not be useful information. That should not be common, though.
* One of the things that SolrCloud will do when it starts is determine whether
any indexes need recovery operations, and proceed to do those operations.
Separate logs indicating when that process starts and when it finishes could be
useful. For the finish event, an elapsed time for all recovery operations
would be essential. And I think it should log the JVM uptime, like the earlier
"startup complete" message.
** Each individual recovery operation should probably log similar events.
was:
Solr should log an event indicating when startup is complete, with a logger
name or a text string that is unique so we can search a log for this event.
The time to log this is when all cores are loaded/started (or inventoried for
lotsOfCores functionality), cloud information is examined and cached as
required, background threads started, etc.
The log message should included JVM uptime and elapsed time since webapp
startup.
If somebody were to hot-deploy Solr into a running container, the JVM uptime
would not be useful information. That should not be common, though.
> Create or improve certain startup log messages
> ----------------------------------------------
>
> Key: SOLR-12173
> URL: https://issues.apache.org/jira/browse/SOLR-12173
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Components: logging
> Affects Versions: 7.3
> Reporter: Shawn Heisey
> Priority: Trivial
>
> Some log events that I think Solr should have when it starts up. All of
> these should have unique logger names or text strings so that a verbose log
> can be searched to find them.
> * An event indicating when startup is complete. The time to log this is
> when all cores are loaded/started (or inventoried for lotsOfCores
> functionality), cloud information is examined and cached as required,
> background threads started, etc.
> ** The log message should included JVM uptime and elapsed time since Solr
> (the webapp) started.
> ** If somebody were to hot-deploy Solr into a running container, the JVM
> uptime would not be useful information. That should not be common, though.
> * One of the things that SolrCloud will do when it starts is determine
> whether any indexes need recovery operations, and proceed to do those
> operations. Separate logs indicating when that process starts and when it
> finishes could be useful. For the finish event, an elapsed time for all
> recovery operations would be essential. And I think it should log the JVM
> uptime, like the earlier "startup complete" message.
> ** Each individual recovery operation should probably log similar events.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
