Turn on some system monitoring.
For that large a difference, look at IO waits or IO reads from disk, whatever
your monitoring system provides.
Also look at CPU, the CPU utilization should be the same on all the nodes. If
some nodes are lower in CPU, then they are blocked by something else.
Mak
On 9/29/23 13:28, rajani m wrote:
What could cause some nodes in a cluster to have high query latency
when compared to the rest?
It is possible that some of the nodes have either handled zero queries
since the last reboot, or that they have handled fewer queries than the
others, and as
Hi、
Thank you, Ufuk.
I understood that it’s not possible to get CSV formatted response from
the terms handler.
I'll use JSON format.
Regards,
mori.
2023年9月30日(土) 0:42 ufuk yılmaz :
>
> When I look at the Solr source code, if I’m understanding correctly, it’s not
> possible to get CSV formatted
In https://github.com/apache/solr/pull/1875 I added a bats test for
replication, and eventually closed it as “not generating enough value over unit
tests”…. Would this test as written have caught your issues?
It would be interesting if you could write a bats test demonstrating the setup
you
Newer information:
Indeed, getting Leader/Follower replication to work was my primary concern: and
that now seems to have been solved.
The reason so many different things were pulled into the thread was that there
did not seem to be any way to know which details were relevant to whatever
myste
Hi Solr Users,
What could cause some nodes in a cluster to have high query latency
when compared to the rest?
It is Solr 9.1.1 cluster of 32 nodes. Cluster has one collection divided
into 32 shards and replication 3, so each node has one Solr server hosting
3 replicas. It is composite route
When I look at the Solr source code, if I’m understanding correctly, it’s not
possible to get CSV formatted response from the terms handler. Because terms
handler puts its data in “terms” key of the map while CSVWriter reads data from
the “response” key only.
//org.apache.solr.handler.component
Thanks.
Yes it seems my version 8.11 doesn´t include that fix to explain more debug
info.
On Fri, 29 Sept 2023 at 15:20, Mikhail Khludnev wrote:
> Hello, Sergio.
> debug=query just displays what ToParentBlockJoinQuery.explain() yields.
> From the beginning ToParentBlockJoinQuery.explain() lacks
On 9/29/23 02:57, John Jackson wrote:
but our requirement NRT so we keep 100 milliseconds.
100 milliseconds is a completely unrealistic target for an automatic
commit interval. On one Solr install I ran where each core had over 25
million documents, 10 seconds was a typical commit time, and
Hello, Sergio.
debug=query just displays what ToParentBlockJoinQuery.explain() yields.
>From the beginning ToParentBlockJoinQuery.explain() lacks functionality
regarding scroreMode that was fixed in Lucene some time ago
https://github.com/apache/lucene/pull/12245. Perhaps the issue you
encounter is
Hi ufuk,
Thank you for your reply.
I'll check & try, then report here.
Regards,
mori.
2023年9月29日(金) 20:35 ufuk yılmaz :
>
> Hello,
>
> Did you try specifying different values for csv.separator or csv.encapsulator
> parameters? Try to choose some unusual characters which can’t exist in terms
As an aside you might look at http://splainer.io which is a cool thing
we built to help you view debug into (it also powers www.quepid.com)
Cheers
Charlie
On 29/09/2023 13:27, Sergio García Maroto wrote:
Hi team,
Lately I have been using some ranking using blockjoin querries on nested
docume
Hi team,
Lately I have been using some ranking using blockjoin querries on nested
documents.
I haved added *score=total *to the queries which actually rings scores to
the parent.
When trying to understand results I see somwhow enabeling debug info only
return first or best match as you can see bel
Hello,
Did you try specifying different values for csv.separator or csv.encapsulator
parameters? Try to choose some unusual characters which can’t exist in terms
themselves.
https://solr.apache.org/guide/solr/latest/query-guide/response-writers.html#csv-response-writer
My first guess is it’s
Hi,
I'm using Solr 9.3.0(Java 11).
About TermsComponent.
When specifying JSON or XML with wt parameter, the expected response
is returned.
But when specifying CSV, only all field names are returned. No data is returned.
What is the solution to this problem?
DATA's
- URL
https:///solr/instan
There seems to be a typo here with the "}"?
>> Yes
It is unusual with 100ms commit time, you risk that commits pile up during
rapid indexing and cause inefficiencies. I'd increase it to at least 1000ms.
>> but our requirement NRT so we keep 100 milliseconds.
Can you reproduce this in an IDLE sys
16 matches
Mail list logo
