Backup strategy with master-slave solr 3.6 servers

We are using solr 3.6 replication with two servers - a master and a slave - and we are currently looking for a way to make clean backups. p>

As the wiki says so , we can use the HTTP command to create a wizard snapshot as follows:http://myMasterHost/solr/replication?command=backup

But we still have a few questions:

  • What is the advantage of a command backupin a classic shell script to copy index files?

  • The team only reserves indexes; is it possible to copy the folder spellchecker? it's necessary?

  • Is it possible to create a snapshot while the application is running, so if there are potential index updates?

  • When we need to restore servers from a backup, what do we need to do on the slave?
    • just copy the snapshot in its index folder and delete the file replication.properties(or not)?
    • to request fetchindex via an HTTP command http://mySlave/solr/replication?command=fetchindex?
    • just delete the sub-index folder to force full replication from the wizard?
+5
source share
1 answer

backup, ReplicationHandler. , , . , Solr. backup.yyyymmddHHMMSS . , .

, , , , , .

, , .

, . , , , , , .

lucene CheckIndex. , , .

, . ( ), , , .

+3

All Articles