Query is taking a time in Solr 6.1.0
Some query is taking time in Solr 6.1.0.
2020-03-12 11:05:36.752 INFO (qtp1239731077-2513155) [c:documents s:shard1
r:core_node1 x:documents] o.a.s.c.S.Request [documents] webapp=/solr
path=/select
params={df=summary&distrib=false&fl=id&shards.purpose=4&start=0&fsv=true&sort=doc_ref+asc,id+desc&fq=&shard.url=s3.test.com:8983/solr/documents|s3r1.test.com:8983/solr/documents&rows=250&version=2&q=(doc_ref:((*n205*)+))+AND+(title:((*Distribution\+Board\+Schedule*)+))+AND+project_id:(2104616)+AND+is_active:true+AND+((isLatest:(true)+AND+isFolderActive:true+AND+isXref:false+AND+-document_type_id:(3+7)+AND+((is_public:true+OR+distribution_list:7249777+OR+folderadmin_list:7249777+OR+author_user_id:7249777)+AND+(((allowedUsers:(7249777)+OR+allowedRoles:(6368666)+OR+combinationUsers:(7249777))+AND+-blockedUsers:(7249777))+OR+(defaultAccess:(true)+AND+-blockedUsers:(7249777)+AND+-blockedRoles:(6368666)+OR+(isLatestRevPrivate:(true)+AND+allowedUsersForPvtRev:(7249777)+AND+-folderadmin_list:(7249777)))&shards.tolerant=true&NOW=1584011129462&isShard=true&wt=javabin}
hits=0 status=0 QTime=7276.
Is there any way to reduce the query execution time(7276 Milli)?
Re: Secure communication between Solr and Zookeeper
This is tracked here https://issues.apache.org/jira/browse/SOLR-7893 - help needed! There are many steps needed, and last time I tried I needed to use Java13 instead of Java11 for some reason. Jan > 12. mar. 2020 kl. 17:23 skrev Kirk Fitzsimons : > > > > Is it possible to secure communication between Solr(8.3.1) and Zookeeper > (3.5.5)? > > We have followed the guide > https://cwiki.apache.org/confluence/display/ZOOKEEPER/ZooKeeper+SSL+User > +Guide > > When we try the below command it hangs and fails to connect to zookeeper. > > solr/server/scripts/cloud-scripts$ ./zkcli.sh -zkhost localhost:2281 -cmd > ls > INFO - 2020-03-12 14:08:41.438; > org.apache.solr.common.cloud.ConnectionManager; Waiting for client to > connect to ZooKeeper > > In the zookeeper logs we can see the following error: > > io.netty.handler.codec.DecoderException: > io.netty.handler.ssl.NotSslRecordException: not an SSL/TLS record > > Kirk Fitzsimons
Should I index the field that use in fq field?
I'm doing a lot of filter query in fq. My search is something like 'q=*:*&fq=..function on a few fields..' . Do I need to only index those field and use FL to get other result or do I need to index everything? -- Sent from: https://lucene.472066.n3.nabble.com/Solr-User-f472068.html
Re: Should I index the field that use in fq field?
Yes, every field you query has to be "indexed" every field you need to be returned in the response has to be "stored" the parameter "fl" can only return fields that are "stored". Other fields given are simply ignored. > On 13 Mar 2020, at 13:15, GTHell wrote: > > I'm doing a lot of filter query in fq. My search is something like > 'q=*:*&fq=..function on a few fields..' . Do I need to only index those > field and use FL to get other result or do I need to index everything? > > > > -- > Sent from: https://lucene.472066.n3.nabble.com/Solr-User-f472068.html
Re: Query is taking a time in Solr 6.1.0
Looks like you have two, maybe three, wildcard/prefix clauses in there.
Consider tokenizing differently so you can optimize the queries to not need
wildcards - thats my first observation and suggestion.
Erik
> On Mar 13, 2020, at 05:56, vishal patel wrote:
>
> Some query is taking time in Solr 6.1.0.
>
> 2020-03-12 11:05:36.752 INFO (qtp1239731077-2513155) [c:documents s:shard1
> r:core_node1 x:documents] o.a.s.c.S.Request [documents] webapp=/solr
> path=/select
> params={df=summary&distrib=false&fl=id&shards.purpose=4&start=0&fsv=true&sort=doc_ref+asc,id+desc&fq=&shard.url=s3.test.com:8983/solr/documents|s3r1.test.com:8983/solr/documents&rows=250&version=2&q=(doc_ref:((*n205*)+))+AND+(title:((*Distribution\+Board\+Schedule*)+))+AND+project_id:(2104616)+AND+is_active:true+AND+((isLatest:(true)+AND+isFolderActive:true+AND+isXref:false+AND+-document_type_id:(3+7)+AND+((is_public:true+OR+distribution_list:7249777+OR+folderadmin_list:7249777+OR+author_user_id:7249777)+AND+(((allowedUsers:(7249777)+OR+allowedRoles:(6368666)+OR+combinationUsers:(7249777))+AND+-blockedUsers:(7249777))+OR+(defaultAccess:(true)+AND+-blockedUsers:(7249777)+AND+-blockedRoles:(6368666)+OR+(isLatestRevPrivate:(true)+AND+allowedUsersForPvtRev:(7249777)+AND+-folderadmin_list:(7249777)))&shards.tolerant=true&NOW=1584011129462&isShard=true&wt=javabin}
> hits=0 status=0 QTime=7276.
>
> Is there any way to reduce the query execution time(7276 Milli)?
Re: Should I index the field that use in fq field?
This isn’t quite right. The fields do _not_ need indexed=true if (and only if) they have docValues=true in order to use function queries. But it’ll consume lots of on-heap resources. In fact, depending on the version of Solr (which you have not told us), it’s even possible to search on a field with docValues=true indexed=false. But it’s extremely inefficient. Which causes you problems when you extrapolate from small data sets to production. Something that works fine with 100K docs blows up when you start putting 10M docs in. So here’s the rule, for efficiency (both run-time and memory): 1> indexed=“true” for anything you search on. I.e. q=field:value or even fq=field:value 2> docValues=“true” for anything you use for function queries, sorting, grouping, or faceting. Of course you can specify both for any given field. Best, Erick > On Mar 13, 2020, at 8:29 AM, Nicolas Franck wrote: > > Yes, > > every field you query has to be "indexed" > > every field you need to be returned in the response has to be "stored" > > the parameter "fl" can only return fields that are "stored". Other fields > given are simply ignored. > > > >> On 13 Mar 2020, at 13:15, GTHell wrote: >> >> I'm doing a lot of filter query in fq. My search is something like >> 'q=*:*&fq=..function on a few fields..' . Do I need to only index those >> field and use FL to get other result or do I need to index everything? >> >> >> >> -- >> Sent from: https://lucene.472066.n3.nabble.com/Solr-User-f472068.html >
Re: Should I index the field that use in fq field?
Does it mean that if the FL field is the result of some function, do I need to only Index those fields right? Or do I need to stored it also? -- Sent from: https://lucene.472066.n3.nabble.com/Solr-User-f472068.html
Re: Should I index the field that use in fq field?
FL has two flavors, neither of them have to do with index=true|false. To return a field in the FL list, specify stored=“true”. To return the results of a _function_, you can do that with “pseudo-fields”, please look in the reference guide. Best, Erick > On Mar 13, 2020, at 8:42 AM, GTHell wrote: > > Does it mean that if the FL field is the result of some function, do I need > to only Index those fields right? Or do I need to stored it also? > > > > -- > Sent from: https://lucene.472066.n3.nabble.com/Solr-User-f472068.html
