On 8/18/2015 7:21 AM, Norgorn wrote:
> SOLR version - 4.10.3
> We have SOLR Cloud cluster, each node has documents only for several
> categories.
> Queries look like "...fq=cat(1 3 89 ...)&..."
> So, only some nodes need to process, others can answer with zero as soon as
> they check "cat".
>
> The
Solr Cloud Document Routing described at
https://cwiki.apache.org/confluence/display/solr/Shards+and+Indexing+Data+in+SolrCloud
allows you to omit hitting certain shards, but they need to be assigned
with the different prefixes beforehand.
Do I get your point right?
On Tue, Aug 18, 2015 at 4:57 PM
Have you tried this with Cache=false?
https://wiki.apache.org/solr/CommonQueryParameters#Caching_of_filters
Because the internal representation of the field value already may be
doing what you want. And the caching of non-repeating filters is what
slowing it down.
I would just do that as a sanity
I'm sorry for being so unclear.
The problem is in speed - while node holds only several cats, it can answer
with "numFound=0", if these cats are missed in query.
It looks like:
node 1 - cats 1,2,3
node 2 - cats 3,4,5
node 3 - cats 50,70
...
Query "q=cat:(1 4)"
QTime per node now is like
node1 -
I am not sure I understand the problem statement. Is it speed? Memory
usage? Something very specific about SolrCloud?
To me it seems the problem is that your 'fq' _are_ getting cached when
you may not want them as the list is different every time. You could
disable that cache.
Or you could try Te
