[
https://issues.apache.org/jira/browse/SOLR-6473?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14119311#comment-14119311
]
Ramkumar Aiyengar commented on SOLR-6473:
-----------------------------------------
If the tlog lookup is a concern, perhaps there could be a parameter to switch
it off with /get?
Ideally search requests shouldn't have to be parsed by the distributing node,
that would help search performance for all requests. It does currently, but
would be good to move in the direction of avoiding it rather that requiring it
for such an optimization.
> Optimize the get-by-id path for search requests
> -----------------------------------------------
>
> Key: SOLR-6473
> URL: https://issues.apache.org/jira/browse/SOLR-6473
> Project: Solr
> Issue Type: Improvement
> Components: SolrCloud
> Reporter: Shalin Shekhar Mangar
> Labels: difficulty-easy, impact-low, workaround-exists
> Fix For: 4.11
>
> Original Estimate: 48h
> Remaining Estimate: 48h
>
> A search request which looks up a doc with just a uniqueKey can be optimised
> to lookup the right shard and send a distrib=false request and avoid the
> overhead of a distributed request.
> The same can also be achieved by the /get handler but it has the (admittedly
> small) overhead of looking up from the transaction log as well.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
