Hi ,
I am plannig to migrate from solr 8.11.3 to 9.5.0 in solrcloud
architecture.
The idea is to create a new line of solrcloud 9.5 where to create the
collections present in solrcloud 8.11.3 and reindex all the content on the
new platform.
My doubt is about the config with which to create the c
Thank you Eric !
Il giorno lun 11 mar 2024 alle ore 17:07 Eric Pugh <
ep...@opensourceconnections.com> ha scritto:
> Definitely. The earliest they would go away is Solr 10…. And by that
> point, the upgrade process should be pretty seamless…
>
>
> > On Mar 11, 2024, at 11:23 AM, Isabella Trev
Hi.
You are correct. Solr is no longer a WAR, it is a standalone app. All attempt
to rip out pieces to make it run elsewehre is unsupported and on your own risk
(though probably possible).
We highly recommend following the documented means of installing Solr into your
environment.
Please treat
Thanks, Dima.
The size of my json was only 8MB due to the small size of documents. I
found the root cause of my problem. It may be useful others so I describe
it short:
1. I used schema.xml
2. I had typo in the field names of my python source code
3. update.autoCreateFields was set to true as defa
On Thu, Feb 29, 2024 at 9:18 AM Henrik B A wrote:
> On Wed, Feb 28, 2024 at 3:14 PM Gus Heck wrote:
>
>> *Here's the full exception:*
>>
>> * org.apache.solr.common.SolrException:
>> org.apache.solr.client.solrj.SolrServerException*
>>
>> You missed the exception message which would be very usef
Im tasked with setting up Apache SOLR to run under Authentication, while not
modifying the existing c# calls to SOLR.
The options available are a basic authentication plugin, and a kerebos plugin
When I try the basic authentication plugin I find that I need to change the
calls to SOLR to includ
Hi Solr Users,
How do you bury results based on subquery such as bq=-(keywords:2023)?
Thank you,
Rajani
It would be interesting to understand the full use case because it feels
like a matching problem rather than a ranking problem (like you don't want
the search result in the first place).
Given that, shooting in the dark, some options come to my mind:
1) you boost anything else (*:* - something)
On 3/13/24 03:06, Isabella Trevisan wrote:
Hi ,
I am plannig to migrate from solr 8.11.3 to 9.5.0 in solrcloud
architecture.
The idea is to create a new line of solrcloud 9.5 where to create the
collections present in solrcloud 8.11.3 and reindex all the content on the
new platform.
My doubt
I recommend checking what has changed between the versions and ensuring
that all the configs used are still valid and supported.
I would also carefully change the constants such as the Lucene match
version (https://sease.io/2021/04/lucenematchversion-in-apache-solr.html).
So, in short, there's no n
Boost factor 0.1 should do it
On Wed, Mar 13, 2024, 9:37 AM rajani m wrote:
> Hi Solr Users,
>
>How do you bury results based on subquery such as bq=-(keywords:2023)?
>
> Thank you,
> Rajani
>
If you know the IDs of the documents and you want to exclude them entirely
(not just sort them last), have a look if the Query Elevation Component
suits your needs.
https://solr.apache.org/guide/solr/latest/query-guide/query-elevation-component.html
Thomas
Op wo 13 mrt 2024 om 17:36 schreef raja
Thank you all.
Robi, yeah boost instead of bq with less than 1 decimals does it, will just
use boost with function query() and see if that helps.
Alessandro, thank you for those pointers, it is a ranking problem, want to
keep the matches as high as possible and at the same time positively boost
g
On 3/11/24 08:37, Isabella Trevisan wrote:
Hi,
I am planning to migrate my solrcloud from 8.11.3 to 9.5.0 .
Our applications user SolJ to work with solr.
I read on the upgrade Notes that I have to use Http2SolrClient instead of
HttpSolrClient .
But on Apache SolR 9.5 reference guide I read tha
ie I was thinking of it like
q=*:*&bq=keywords:2023^0.01&bq=-keywords:2023^2
but I'm thinking your way was sufficient. Did that not work?
On Wed, Mar 13, 2024 at 10:52 AM Robi Petersen wrote:
> Boost factor 0.1 should do it
>
> On Wed, Mar 13, 2024, 9:37 AM rajani m wrote:
>
>> Hi Sol
oh missed your response NM lol
On Wed, Mar 13, 2024 at 9:31 PM Robi Petersen wrote:
> ie I was thinking of it like
> q=*:*&bq=keywords:2023^0.01&bq=-keywords:2023^2
> but I'm thinking your way was sufficient. Did that not work?
>
>
> On Wed, Mar 13, 2024 at 10:52 AM Robi Petersen wrote:
>
16 matches
Mail list logo