[
https://issues.apache.org/jira/browse/SOLR-1216?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12719078#action_12719078
]
Noble Paul commented on SOLR-1216:
----------------------------------
bq.Lets also clearup the doc for this stuff as part of the issue....
yes the wiki has to be cleaned up too
bq.Are backups made in the same format as the scripts method
The format is same (same dir name format). The mechanism is different. scripts
use hard links , this uses copying
bq.Is it possible to replicate a backup or just the live index?
nope. only live index can be transferred. The backup can be made the live index
and then the user can replicate it. There is a file called index.properties in
the dataDir (only created by replicationhandler. but users can create them too
). It can take a property index=<dir-name> which tells solr core where to load
the index from. users can edit that and do a reload core for the new backup to
be used as the index.
bq.what happens when you shutdown during a replication - does Solr wait? Abort?
If its aborts, are the temp files cleaned up later?
replication is aborted. temp files will remain as is. They will not be cleaned
up automatically. But presence of those temp files will not have any impact of
future replication.
bq.replicate and polling both to default to on right?
yes
bq.and the replicate setting just stops any slaves hitting a master with
replicate=off from syncing?
the slaves will keep polling even if the master has disabled replication . But
the master will respond with "no changes" .This will ensure that when the
replication is re-enabled from master , the slaves will be able to resume
replication
> disambiguate the replication command names
> ------------------------------------------
>
> Key: SOLR-1216
> URL: https://issues.apache.org/jira/browse/SOLR-1216
> Project: Solr
> Issue Type: Improvement
> Components: replication (java)
> Reporter: Noble Paul
> Assignee: Noble Paul
> Fix For: 1.4
>
>
> There is a lot of confusion in the naming of various commands such as
> snappull, snapshot etc. This is a vestige of the script based replication we
> currently have. The commands can be renamed to make more sense
> * 'snappull' to be renamed to 'sync'
> * 'snapshot' to be renamed to 'backup'
> thoughts?
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
