Hello,

We've got a big index with millions of documents describing a relationship 
between entities.  One entity and it's many friends.  We score them in a 
few different ways and later on query and show paged results by those 
scores.

The question is how were warmers supposed to be used most effectively.  

Currently we have segmented a group of users that we want always want warm 
in the cache, but it seems like after a while and a heavy write load the 
warmer does not keep those users well.. 'warm'.  The initial load still 
takes a while.

Also another thing we're not sure about it, is whether we should create a 
big warmer query that would order all friends by the score or create the 
same queries (thus creating many smaller warmers) as we would use on the 
web page.

Appreciate your thoughts and more insight into how those warmers work and 
how we can achieve a more smooth query latency result.

Thanks much!
Daniel

-- 
You received this message because you are subscribed to the Google Groups 
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/elasticsearch/66c5c4b0-dd9a-42df-83c4-7b62c0746449%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to