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 "Markus.Mirsberger" <ma...@gmx.de> on 2013/02/10 15:38:30 UTC

Problem with shard naming in Solr 4.1

Hi,

I upgraded today from a 4.1 nightly build to the 4.1 realease.
Now I noticed that the naming of collection shards changed.

When I create a collection an identifier for the shard and the replica 
will be added automatically?
Can I turn this off somehow or is there a way to retrieve the corename 
of the collection shard on a specific server?

If not this means I can no longer access one shard of the collection 
through a HttpSolrServer in SolrJ because I can not know which ID a 
shard have on a server.

Some operations are impossible to do with the SolrCloud....e.g. 
iterating through and updating all documents of a collection with 
millions of documents because every shard needs to send all previous 
documents to the leader too when iterating.
Even when I use the shards parameter to access only one specific shard 
all documents will be sent to the leader first before I get them.

So I need the possibility to access shards of collections directly with 
a HttpSolrServer.
Any ideas how I can do this? Maybe I just missed another possibility.




Regards,
Markus