You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-dev@lucene.apache.org by "Bill Au (JIRA)" <ji...@apache.org> on 2010/01/26 23:06:34 UTC
[jira] Created: (SOLR-1734) Add pid file to snappuller to skip
script overruns, and recover from failure
Add pid file to snappuller to skip script overruns, and recover from failure
----------------------------------------------------------------------------
Key: SOLR-1734
URL: https://issues.apache.org/jira/browse/SOLR-1734
Project: Solr
Issue Type: Improvement
Components: replication (scripts)
Affects Versions: 1.4
Reporter: Bill Au
Assignee: Bill Au
Priority: Minor
The pid file will allow snappuller to be run as fast as possible without overruns. Also it will recover from a last failed run should an older snappuller process no longer be running. The same has already been done to snapinstaller in SOLR-990. Overlapping snappuller could cause replication traffic to saturate the network if a large Solr index is being replicated to a large number of clients.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (SOLR-1734) Add pid file to snappuller to skip
script overruns, and recover from failure
Posted by "Bill Au (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/SOLR-1734?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Bill Au updated SOLR-1734:
--------------------------
Attachment: SOLR-1734-2.patch
> Add pid file to snappuller to skip script overruns, and recover from failure
> ----------------------------------------------------------------------------
>
> Key: SOLR-1734
> URL: https://issues.apache.org/jira/browse/SOLR-1734
> Project: Solr
> Issue Type: Improvement
> Components: replication (scripts)
> Affects Versions: 1.4
> Reporter: Bill Au
> Assignee: Bill Au
> Priority: Minor
> Attachments: SOLR-1734-2.patch, SOLR-1734.patch
>
>
> The pid file will allow snappuller to be run as fast as possible without overruns. Also it will recover from a last failed run should an older snappuller process no longer be running. The same has already been done to snapinstaller in SOLR-990. Overlapping snappuller could cause replication traffic to saturate the network if a large Solr index is being replicated to a large number of clients.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (SOLR-1734) Add pid file to snappuller to skip
script overruns, and recover from failure
Posted by "Bill Au (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/SOLR-1734?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Bill Au updated SOLR-1734:
--------------------------
Attachment: SOLR-1734.patch
I am reusing the code from SOLR-990 which adds the same feature to snapinstaller. I have added a "-f" command line argument to force the snappuller to run even if one is already running. That will be useful if network capacity is not an issue.
> Add pid file to snappuller to skip script overruns, and recover from failure
> ----------------------------------------------------------------------------
>
> Key: SOLR-1734
> URL: https://issues.apache.org/jira/browse/SOLR-1734
> Project: Solr
> Issue Type: Improvement
> Components: replication (scripts)
> Affects Versions: 1.4
> Reporter: Bill Au
> Assignee: Bill Au
> Priority: Minor
> Attachments: SOLR-1734.patch
>
>
> The pid file will allow snappuller to be run as fast as possible without overruns. Also it will recover from a last failed run should an older snappuller process no longer be running. The same has already been done to snapinstaller in SOLR-990. Overlapping snappuller could cause replication traffic to saturate the network if a large Solr index is being replicated to a large number of clients.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (SOLR-1734) Add pid file to snappuller to skip
script overruns, and recover from failure
Posted by "Bill Au (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/SOLR-1734?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Bill Au updated SOLR-1734:
--------------------------
Attachment: SOLR-1734-2.patch
The original patch has a slight race condition within the time window between testing for the presence and creation of the lock file. Since snappuller is a bash script, the revised patch uses the noclobber option to prevent that.
> Add pid file to snappuller to skip script overruns, and recover from failure
> ----------------------------------------------------------------------------
>
> Key: SOLR-1734
> URL: https://issues.apache.org/jira/browse/SOLR-1734
> Project: Solr
> Issue Type: Improvement
> Components: replication (scripts)
> Affects Versions: 1.4
> Reporter: Bill Au
> Assignee: Bill Au
> Priority: Minor
> Attachments: SOLR-1734-2.patch, SOLR-1734.patch
>
>
> The pid file will allow snappuller to be run as fast as possible without overruns. Also it will recover from a last failed run should an older snappuller process no longer be running. The same has already been done to snapinstaller in SOLR-990. Overlapping snappuller could cause replication traffic to saturate the network if a large Solr index is being replicated to a large number of clients.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (SOLR-1734) Add pid file to snappuller to skip
script overruns, and recover from failure
Posted by "Bill Au (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/SOLR-1734?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Bill Au updated SOLR-1734:
--------------------------
Attachment: (was: SOLR-1734-2.patch)
> Add pid file to snappuller to skip script overruns, and recover from failure
> ----------------------------------------------------------------------------
>
> Key: SOLR-1734
> URL: https://issues.apache.org/jira/browse/SOLR-1734
> Project: Solr
> Issue Type: Improvement
> Components: replication (scripts)
> Affects Versions: 1.4
> Reporter: Bill Au
> Assignee: Bill Au
> Priority: Minor
> Attachments: SOLR-1734-2.patch, SOLR-1734.patch
>
>
> The pid file will allow snappuller to be run as fast as possible without overruns. Also it will recover from a last failed run should an older snappuller process no longer be running. The same has already been done to snapinstaller in SOLR-990. Overlapping snappuller could cause replication traffic to saturate the network if a large Solr index is being replicated to a large number of clients.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.