You are viewing a plain text version of this content. The canonical link for it is here.
Posted to solr-user@lucene.apache.org by garlandkr <ga...@gmail.com> on 2012/06/27 23:30:48 UTC

Index Snapshot for 3.3?

How can I get a snapshot of the index in SOLR 3.x? 

I am currently taking EBS (Amazon) snapshots of the volume where the data is
form one machine and creating new volumes form that snapshot. When the
service starts it still runs through an indexing process that takes forever.
Is there a way to get a snapshot of the index and use that without having to
re-index on the new system?

--
View this message in context: http://lucene.472066.n3.nabble.com/Index-Snapshot-for-3-3-tp3991684.html
Sent from the Solr - User mailing list archive at Nabble.com.

Re: Index Snapshot for 3.3?

Posted by Upayavira <uv...@odoko.co.uk>.
Use once off replication, or, if you prefer, on unix do

 cp -lr <your-index-dir> <your-backup-dir>

at a time you know a commit isn't happening.

You'll have a clone of the index you can ship to another host. Remember
to delete your backup when done.

This uses the fact that files in a Lucene index never change, they're
only ever deleted and new ones added. Thus, using hard links to clone an
index works well.

Upayavira

On Wed, Jun 27, 2012, at 02:30 PM, garlandkr wrote:
> How can I get a snapshot of the index in SOLR 3.x? 
> 
> I am currently taking EBS (Amazon) snapshots of the volume where the data
> is
> form one machine and creating new volumes form that snapshot. When the
> service starts it still runs through an indexing process that takes
> forever.
> Is there a way to get a snapshot of the index and use that without having
> to
> re-index on the new system?
> 
> --
> View this message in context:
> http://lucene.472066.n3.nabble.com/Index-Snapshot-for-3-3-tp3991684.html
> Sent from the Solr - User mailing list archive at Nabble.com.