[ 
https://issues.apache.org/jira/browse/SOLR-4816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13656232#comment-13656232
 ] 

Mark Miller edited comment on SOLR-4816 at 5/13/13 6:37 PM:
------------------------------------------------------------

bq.  It also looks like it won't work with batches because it checks the first 
document id only.

Yes, see the comments in the issue:

"The patch is limited, but a start: right now it's just for String and Integer 
Id's and it only acts upon the first document or deleteby id (favoring 
document) - if you use the bulk methods they are all sent along to the leader 
of the first id."

bq.  I like the idea of a nice clean separate method for this.

I still think it needs to be the default method, and we simply should support 
batching. The other limitation around types is no longer a concern now that 
Yonik changed the hashing.

I don't know that we need to build on the patch in SOLR-3154, it was a very 
quick experiment, but I think that is the right implementation direction. The 
smart client should simply work optimally without having to use alternate 
methods IMO.
                
      was (Author: [email protected]):
    bq.  It also looks like it won't work with batches because it checks the 
first document id only.

Yes, see the comments in the issue:

"The patch is limited, but a start: right now it's just for String and Integer 
Id's and it only acts upon the first document or deleteby id (favoring 
document) - if you use the bulk methods they are all sent along to the leader 
of the first id."

bq.  I like the idea of a nice clean separate method for this.

I still think it needs to be the default method, and we simply should support 
batching. The other limitation around types is no longer a concern now that 
Yonik changing the hashing.

I don't know that we need to build on the patch in SOLR-3154, it was a very 
quick experiment, but I think that is the right implementation direction. The 
smart client should simply work optimally without having to use alternate 
methods IMO.
                  
> Add method to CloudSolrServer to send updates to the correct shard
> ------------------------------------------------------------------
>
>                 Key: SOLR-4816
>                 URL: https://issues.apache.org/jira/browse/SOLR-4816
>             Project: Solr
>          Issue Type: Improvement
>          Components: SolrCloud
>    Affects Versions: 4.3
>            Reporter: Joel Bernstein
>            Priority: Minor
>         Attachments: SOLR-4816.patch
>
>
> This issue adds a "directUpdate" method to CloudSolrServer which routes 
> update requests to the correct shard. This would be a nice feature to have to 
> eliminate the document routing overhead on the Solr servers.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to