Thanks, David, Shawn, Jagdish
Help and suggestions are really appreciated.
Regards,
Lucky Sharma
On Thu, Jun 27, 2019 at 12:50 AM Shawn Heisey wrote:
>
> On 6/26/2019 12:56 PM, Lucky Sharma wrote:
> > @Shawn: Sorry I forgot to mention the corpus size: the corpus size is
> > around 3 million
On 6/26/2019 12:56 PM, Lucky Sharma wrote:
@Shawn: Sorry I forgot to mention the corpus size: the corpus size is
around 3 million docs, where we need to query for 1500 docs and run
aggregations, sorting, search on them.
Assuming the documents aren't HUGE, that sounds like something Solr
Then term query parser is best way to do that.
You can check below link for performance detail.
http://yonik.com/solr-terms-query/
n Thu, 27 Jun, 2019, 12:31 AM Lucky Sharma, wrote:
> Thanks, Jagdish
> But what if we need to perform search and filtering on those 1.5k doc
> ids results, also
yeah there is a performance hit but that is expected. in my scenario i
pass sometimes a few thousand using this method, but i pre-process my
results since its a set. you will not have any issues if you are using
POST with the uri length.
On Wed, Jun 26, 2019 at 3:02 PM Lucky Sharma wrote:
>
Thanks, Jagdish
But what if we need to perform search and filtering on those 1.5k doc
ids results, also for URI error, we can go with the POST approach,
and what if the data is not sharded.
Regards,
Lucky Sharma
On Thu, Jun 27, 2019 at 12:28 AM jai dutt wrote:
>
> 1. No Solr is not for id
1. No Solr is not for id search. rdms a better option.
2. Yes correct it going to impact query performance. And you may got
large uri error.
3 ya you can pass ids internally by writing any custom parser.or divide
data into different shard.
On Thu, 27 Jun, 2019, 12:01 AM Lucky Sharma,
@Shawn: Sorry I forgot to mention the corpus size: the corpus size is
around 3 million docs, where we need to query for 1500 docs and run
aggregations, sorting, search on them.
@David: But will that not be a performance hit (resource incentive)?
since it will have that many terms to search upon,
you can use the !terms operator and send them separated by a comma:
{!terms f=id}id1,id2,..id1499,id1500
and run facets normally
On Wed, Jun 26, 2019 at 2:31 PM Lucky Sharma wrote:
> Hi all,
>
> What we are doing is, we will be having a set of unique Ids of solr
> document at max 1500,
On 6/26/2019 12:31 PM, Lucky Sharma wrote:
What we are doing is, we will be having a set of unique Ids of solr
document at max 1500, we need to run faceting and sorting among them.
there is no direct search involved.
It's a head-on search since we already know the document unique keys
