Re: Understanding the MoreLikeThis Handler
This has been resolved. The issue was indeed the mlt.mintf value. -- View this message in context: http://lucene.472066.n3.nabble.com/Understanding-the-MoreLikeThis-Handler-tp4213279p4213399.html Sent from the Solr - User mailing list archive at Nabble.com.
Index time field boost with solrj?
Hello! Can I boost fields at index time with the solrj client? If not what do you recommend? Feels like it would be expensive to always do it at query time? Cheers
Re: Solr Exact match boost Reduce the results
Hi Jack,
2015-06-23 6:27 GMT+01:00 JACK :
> Hi Alessandro Benedetti,
>
> I have changed the query like this.
>
>
> /select?q=("dell+laptop"~13+OR+"dell+laptop")&df=product_name&defType=edismax&qf=product_name^0.001+dummy_product_name^2&fl=product_name&wt=json&indent=true&debug=true
>
Pardon me, but I find this query awful. In my opinion is not necessary at
all.
I can't take a deep look to your debug right now, but I suggest you to use
this free tool ( http://splainer.io ) to read it easily .
My suggestions remain :
1) carefully take care of your analysis
2) carefully take care of the schema ( field attributes)
3) carefully take care of the handler config
Cheers
>
> The corresponding results also given in the below link.Now am getting exact
> match first.
>
> http://pastebin.com/rAYrFiB8
>
>
> Now the problem is Look at the 8th result
> "product_name":"Dell Inspiron 15R 15.6-inch Laptop without Laptop Bag by
> Dell", its not a relevant result
> can you check How this happens , check the scores.
>
>
>
> --
> View this message in context:
> http://lucene.472066.n3.nabble.com/Solr-Exact-match-boost-Reduce-the-results-tp4211352p4213382.html
> Sent from the Solr - User mailing list archive at Nabble.com.
>
--
--
Benedetti Alessandro
Visiting card : http://about.me/alessandro_benedetti
"Tyger, tyger burning bright
In the forests of the night,
What immortal hand or eye
Could frame thy fearful symmetry?"
William Blake - Songs of Experience -1794 England
Re: Understanding the MoreLikeThis Handler
You are welcome ;) 2015-06-23 9:27 GMT+01:00 Sreekant Sreedharan : > This has been resolved. The issue was indeed the mlt.mintf value. > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/Understanding-the-MoreLikeThis-Handler-tp4213279p4213399.html > Sent from the Solr - User mailing list archive at Nabble.com. > -- -- Benedetti Alessandro Visiting card : http://about.me/alessandro_benedetti "Tyger, tyger burning bright In the forests of the night, What immortal hand or eye Could frame thy fearful symmetry?" William Blake - Songs of Experience -1794 England
Question regarding getting back results from middle of word
Hello! if I have a field like "onetwothreefour" I want users to find that document with a query for "two". This seems to work when I use I got that one off google. The problem is I really want a search for "one" to be regarded as a better hit but seems the fragments created are like an equal word, or am I missing something? I also don't want those fragments to be a part of the spelling index... Feels like I am missing something on how to best configure wildcards in a word yet not affecting my index. Cheers
Re: Solr Exact match boost Reduce the results
Hi Alessandro Benedetti, Can you check what happens to the below given product_name , just check its score "product_name":"LAPTOP BATTERY DELL Inspiron 6400 1501 E1505 RD859 UD267 XU937" How is this product comes , ? -- View this message in context: http://lucene.472066.n3.nabble.com/Solr-Exact-match-boost-Reduce-the-results-tp4211352p4213417.html Sent from the Solr - User mailing list archive at Nabble.com.
Implicit Router Configurations
We have a use case where documents are indexed in shards according to a specific field (shard per user), and the number of shards is unknown when creating the collection. For that purpose we use the implicit router and define router.field=user. >From what we've seen, the only way to define an implicit router is during the collection creation. Moreover, the router definitions (router.name and router.field) are kept only in clusterstate.json and not in any solr configuration file on disk. In some cases solr state becomes inconsistent and we need to delete the configs from ZK and restart the solr server. The behavior we see is the new clusterstate.json generated by solr on startup has the default router.name=compositeId, which is not what we defined during creation. Are we missing something? Is there a place to configure the implicit router on disk such that it will be persistent? Thanks, Arnon
copyField based on value of another field
Hi folks, is it possible to copyField only if another field has a certain value? e.g. copyField 'dc.subject' to 'image_suggestions' only if rdf http://www.nsdl.org/ontologies/relationships#isInImageBank is true thanks, Alistair -- mov eax,1 mov ebx,0 int 80h
Costume searchHandlers in a multi core envirment
Hi, I have a search handler I wrote that runs a sub query so that the main query I send to the SearchHandler is extended. This was working very nice till I changed my env to a be multi sharded. Now my searcher is always opened to the same core and not to the core that the shard asks for. For example: I have core1 & core2. my http request : localhost:8983/solr/core1/myselect?q=text%3Aabc& shards=http://localhost:8983/solr/core1%20http://localhost:8983/solr/core2 I get to myselect handler twice but the searcher opens to core1 in both cases never to core2. Any ideas? -- View this message in context: http://lucene.472066.n3.nabble.com/Costume-searchHandlers-in-a-multi-core-envirment-tp4213436.html Sent from the Solr - User mailing list archive at Nabble.com.
Re: Question regarding getting back results from middle of word
It sounds the classic XY problem , can you explain us a little bit better your problem ? Why you have such strange field content, how do you produce it ? Can this be solved with an analysis ad hoc for your language ? It sounds to me as a tokenization problem, and you are not going to solve it using edge grams … Please give us more details, Cheers 2015-06-23 11:41 GMT+01:00 Karl Kildén : > Hello! > > if I have a field like "onetwothreefour" I want users to find that document > with a query for "two". This seems to work when I use > > > maxGramSize="25" /> > > > I got that one off google. The problem is I really want a search for "one" > to be regarded as a better hit but seems the fragments created are like an > equal word, or am I missing something? I also don't want those fragments to > be a part of the spelling index... > > Feels like I am missing something on how to best configure wildcards in a > word yet not affecting my index. > > Cheers > -- -- Benedetti Alessandro Visiting card : http://about.me/alessandro_benedetti "Tyger, tyger burning bright In the forests of the night, What immortal hand or eye Could frame thy fearful symmetry?" William Blake - Songs of Experience -1794 England
Re: copyField based on value of another field
You should work at the UpdateProcessor level : https://wiki.apache.org/solr/UpdateRequestProcessor#Implementing_a_conditional_copyField This should give you some hint. Cheers 2015-06-23 13:45 GMT+01:00 Alistair Young : > Hi folks, > > is it possible to copyField only if another field has a certain value? e.g. > > copyField 'dc.subject' to 'image_suggestions' only if rdf > http://www.nsdl.org/ontologies/relationships#isInImageBank is true > > thanks, > > Alistair > > -- > mov eax,1 > mov ebx,0 > int 80h > -- -- Benedetti Alessandro Visiting card : http://about.me/alessandro_benedetti "Tyger, tyger burning bright In the forests of the night, What immortal hand or eye Could frame thy fearful symmetry?" William Blake - Songs of Experience -1794 England
Re: Index time field boost with solrj?
On 6/23/2015 4:42 AM, Karl Kildén wrote: > Hello! Can I boost fields at index time with the solrj client? If not what > do you recommend? Feels like it would be expensive to always do it at query > time? Yes. On the SolrInputDocument class, there are addField and setField variants that include a float parameter for boost: http://lucene.apache.org/solr/5_2_0/solr-solrj/org/apache/solr/common/SolrInputDocument.html Thanks, Shawn
Re: Costume searchHandlers in a multi core envirment
Let's start from this : "I have a search handler I wrote that runs a sub query so that the main query I send to the SearchHandler is extended." What is the problem you are searching to solve ? Why a classic filter query or re-ranking query is not ok for you ? Can you give us some indication of that ? Cheers 2015-06-23 13:48 GMT+01:00 1234go1 <1234...@gmail.com>: > Hi, > I have a search handler I wrote that runs a sub query so that the main > query > I send to the SearchHandler is extended. > This was working very nice till I changed my env to a be multi sharded. > Now my searcher is always opened to the same core and not to the core that > the shard asks for. > For example: > I have core1 & core2. > my http request : > localhost:8983/solr/core1/myselect?q=text%3Aabc& > shards=http://localhost:8983/solr/core1%20http://localhost:8983/solr/core2 > > I get to myselect handler twice but the searcher opens to core1 in both > cases never to core2. > Any ideas? > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/Costume-searchHandlers-in-a-multi-core-envirment-tp4213436.html > Sent from the Solr - User mailing list archive at Nabble.com. > -- -- Benedetti Alessandro Visiting card : http://about.me/alessandro_benedetti "Tyger, tyger burning bright In the forests of the night, What immortal hand or eye Could frame thy fearful symmetry?" William Blake - Songs of Experience -1794 England
Solr returning weird results
I am trying to set up Solr for my data, I have configured all the things and the search seems to be working fine but some times I am getting weird results. For example if I am searching for a term 'vicryl ctx 2-0 70' in the results I have 1. VICRYL VIOLETT GEFL | 2-0 | 1 x CTX PLUS | 1 x 70 | - | 36 2. VICRYL PLUS VIOLETT GEFL | 2-0 | 1 x CTX PLUS | 1 x 70 | - | 36 3. SPONGOSTAN* Film Gelatinehämostyptikum, 200 x 70 x 0,5 mm 4. SPONGOSTAN* Standard Gelatinehämostyptikum, 70 x 50 x 10 mm 5. SPONOGSTAN* Spezial Gelatinehämostyptikum, 70 x 50 x 1 mm 6. VICRYL VIOL GEFL VIOLETT GEFL | 2-0 | 1 x SH1 PLUS | 1 x 70 | - | 12 I don't understand why I am getting results no.4 and no.5 before no.6. I looked at analysis and I found nothing wrong with that... Below is my schema.xml Any help or suggestions will be much appreciated..Thanks -- View this message in context: http://lucene.472066.n3.nabble.com/Solr-returning-weird-results-tp4213467.html Sent from the Solr - User mailing list archive at Nabble.com.
Re: Nested objects and map in Solr
I found the answer to my problems:
1. If you have sub-objects add "content_type": , to the parent
object, e.g.
{
“name_s” : ”mine”,
"content_type": “surnames”,
values {
“surname1” : “first”,
“surname2” : “second"
}
}
2. Solr is very picky and will not index an object if it is not well construct
(case of Json) or any of the object's fields don’t comply with their own type.
Particularly I found it with the date/time, one of the valid formats is
-MM-ddTHH:mm:ssZ, e.g. 2015-06-23T13:30:20Z
Hope this help someone else
> On 22 Jun 2015, at 18:45, Amadeo Asco wrote:
>
> Thank you for your reply.
>
> Some of my objects are not fully indexed. I have 121 objects with sub-objects
> and when querying with “*:*” only 81 are found. The objects are in Json
> format. I use Solr from within Riak.
>
>
> On 22 Jun 2015, at 17:31, Alessandro Benedetti
> mailto:benedetti.ale...@gmail.com>> wrote:
>
> Join is not only query time related.
> Join is the Solr approach to nested documents.
> Block join and Query time join are 2 different strategies with different
> pros and cons .
> Basically :
>
> Block Join :
> - force you to follow a specific indexing strategy
> - if you want to update a child you need to update the whole block ( parent
> +children)
> + 4-5 times faster then query time join
>
> Query Time Join
> + you can index the data how you want, even in different cores
> + you can update single children
> - it is "slow"
>
> What do you mean with "strange" indexes ?
>
> Cheers
>
> 2015-06-22 11:28 GMT+01:00 Amadeo Asco
> mailto:a...@trifork.com>>:
>
> I have looked at the join but the problem isn’t with the query but on
> indexing sub-objects (in some places called also nested objects). The best
> documentation I have seen is http://yonik.com/solr-4-8-features/
> but I am still getting very estrange indexes when I play with this
> approach and others.
>
> Cheers
>
> On 22 Jun 2015, at 11:23, Alessandro Benedetti
> mailto:benedetti.ale...@gmail.com>
> > wrote:
>
> I would suggest you to take a look to the Solr Join ( block and query time
> join) .
> This is what you are looking for.
> Anyway, this request points out that the documentation is not good enough
> to address people to Nested Objects problems.
> Maybe this should highlight the need of improving that part of the
> documentation.
>
> Cheers
>
> 2015-06-22 9:57 GMT+01:00 Amadeo Asco
> mailto:a...@trifork.com> a...@trifork.com>>:
>
> Dear all,
>
> I am starting with Solr and I cannot index map(key,values) or nested
> objects in Solr. I have not been able to find a satisfactory answer online.
> Please could anyone let me know where to find the full documentation of how
> to index nested objects in Solr? Also how to index and reference in the
> search melees in a map?
>
> Example Json:
> {"name_s":"Lion-o", "age_i":30, "leader_b":true, "data":{"name":"me",
> "surname":"other"}, "values_ss":["red", "blue”]}
> {"name_s":"Cheetara", "age_i":28, "leader_b":false, "content_type":
> "parentDocument",
> "_childDocuments_":[{"name":"1"},{"name":"2"},{"name":"3"}]}
>
> What about with:
> {"name_s":"Lion-o", "age_i":30, "leader_b":true, "data":{"name”:[“me”],
> "surname”:[“other”, “next"]}, "values_ss":["red", "blue”]}
>
> I look forward to hearing form you soon.
>
> Thank you in advance,
> Amadeo
>
>
>
>
> --
> --
>
> Benedetti Alessandro
> Visiting card : http://about.me/alessandro_benedetti
>
> "Tyger, tyger burning bright
> In the forests of the night,
> What immortal hand or eye
> Could frame thy fearful symmetry?"
>
> William Blake - Songs of Experience -1794 England
>
>
>
>
> --
> --
>
> Benedetti Alessandro
> Visiting card : http://about.me/alessandro_benedetti
>
> "Tyger, tyger burning bright
> In the forests of the night,
> What immortal hand or eye
> Could frame thy fearful symmetry?"
>
> William Blake - Songs of Experience -1794 England
>
Re: Solr returning weird results
Sent from my iPhone > On 23 Jun 2015, at 16:37, Freakheart wrote: > > I am trying to set up Solr for my data, I have configured all the things and > the search seems to be working fine but some times I am getting weird > results. > > For example if I am searching for a term 'vicryl ctx 2-0 70' in the results > I have > > 1. VICRYL VIOLETT GEFL | 2-0 | 1 x CTX PLUS | 1 x 70 | - | 36 > 2. VICRYL PLUS VIOLETT GEFL | 2-0 | 1 x CTX PLUS | 1 x 70 | - | 36 > 3. SPONGOSTAN* Film Gelatinehämostyptikum, 200 x 70 x 0,5 mm > 4. SPONGOSTAN* Standard Gelatinehämostyptikum, 70 x 50 x 10 mm > 5. SPONOGSTAN* Spezial Gelatinehämostyptikum, 70 x 50 x 1 mm > 6. VICRYL VIOL GEFL VIOLETT GEFL | 2-0 | 1 x SH1 PLUS | 1 x 70 | - | 12 > > I don't understand why I am getting results no.4 and no.5 before no.6. I > looked at analysis and I found nothing wrong with that... > > Below is my schema.xml > > positionIncrementGap="100"> > > > > > > > > > > > positionIncrementGap="100"> > > > > > > > > > > > Any help or suggestions will be much appreciated..Thanks > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/Solr-returning-weird-results-tp4213467.html > Sent from the Solr - User mailing list archive at Nabble.com.
Re: Solr returning weird results
In this situations I would suggest you to use the DebugQuery=true. To simplify the understanding you can also use : splainer.io . Really nice tool to quickly identify why a document is there. Cheers 2015-06-23 15:37 GMT+01:00 Freakheart : > I am trying to set up Solr for my data, I have configured all the things > and > the search seems to be working fine but some times I am getting weird > results. > > For example if I am searching for a term 'vicryl ctx 2-0 70' in the results > I have > > 1. VICRYL VIOLETT GEFL | 2-0 | 1 x CTX PLUS | 1 x 70 | - | 36 > 2. VICRYL PLUS VIOLETT GEFL | 2-0 | 1 x CTX PLUS | 1 x 70 | - | 36 > 3. SPONGOSTAN* Film Gelatinehämostyptikum, 200 x 70 x 0,5 mm > 4. SPONGOSTAN* Standard Gelatinehämostyptikum, 70 x 50 x 10 mm > 5. SPONOGSTAN* Spezial Gelatinehämostyptikum, 70 x 50 x 1 mm > 6. VICRYL VIOL GEFL VIOLETT GEFL | 2-0 | 1 x SH1 PLUS | 1 x 70 | - | 12 > > I don't understand why I am getting results no.4 and no.5 before no.6. I > looked at analysis and I found nothing wrong with that... > > Below is my schema.xml > > positionIncrementGap="100"> > > > > > > > > > > > positionIncrementGap="100"> > > > > > > > > > > > Any help or suggestions will be much appreciated..Thanks > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/Solr-returning-weird-results-tp4213467.html > Sent from the Solr - User mailing list archive at Nabble.com. > -- -- Benedetti Alessandro Visiting card : http://about.me/alessandro_benedetti "Tyger, tyger burning bright In the forests of the night, What immortal hand or eye Could frame thy fearful symmetry?" William Blake - Songs of Experience -1794 England
"More" facets or "Less" facets in the Velocity UI
Hi Is there a way to add a link "More" (or "Less") under the list of facet's values, in aim to display more facets values? Thanks! Benjamin
Upgrading SolrCloud 4.8 to 4.8.1
Hi All, I'm upgrading a SolrCluster from 4.8 to 4.8.1. Looking at collections configuration there are few solrconfig.xml that have LUCENE_48 others have: 4.8 Now I'm updating everything to: 4.8.1 Given that after the change I have to reindex all the collections. In your experience, is there any other counter-indication? Best regards, Vincenzo -- Vincenzo D'Amore email: v.dam...@gmail.com skype: free.dev mobile: +39 349 8513251
Re: Upgrading SolrCloud 4.8 to 4.8.1
On 6/23/2015 9:37 AM, Vincenzo D'Amore wrote: > I'm upgrading a SolrCluster from 4.8 to 4.8.1. > Looking at collections configuration there are few solrconfig.xml that have > > LUCENE_48 > > others have: > > 4.8 > > Now I'm updating everything to: > > 4.8.1 Changing your luceneMatchVersion is unnecessary. We try to never make changes in the API or Solr behavior in a bugfix release, unless it's absolutely necessary to fix a bug. To my knowledge, changes in API and general behavior have NOT been required by any recent bugfix release. This means that a 4.8.1 match version will produce exactly the same results as a 4.8 match version. Changes in the index format are also avoided, so you will not need to reindex. The luceneMatchVersion code is capable of dealing with X.Y versions as well as LUCENE_XY ... Starting with 4.9, it also understands LUCENE_X_Y ... which was a required addition because version 4.10 wouldn't have worked with the LUCENE_XY notation. Thanks, Shawn
Re: Upgrading SolrCloud 4.8 to 4.8.1
Thank you very much Shawn, very interesting. Is there any chance to find this behaviour explained in the wiki? On Tue, Jun 23, 2015 at 6:02 PM, Shawn Heisey wrote: > On 6/23/2015 9:37 AM, Vincenzo D'Amore wrote: > > I'm upgrading a SolrCluster from 4.8 to 4.8.1. > > Looking at collections configuration there are few solrconfig.xml that > have > > > > LUCENE_48 > > > > others have: > > > > 4.8 > > > > Now I'm updating everything to: > > > > 4.8.1 > > Changing your luceneMatchVersion is unnecessary. We try to never make > changes in the API or Solr behavior in a bugfix release, unless it's > absolutely necessary to fix a bug. To my knowledge, changes in API and > general behavior have NOT been required by any recent bugfix release. > This means that a 4.8.1 match version will produce exactly the same > results as a 4.8 match version. Changes in the index format are also > avoided, so you will not need to reindex. > > The luceneMatchVersion code is capable of dealing with X.Y versions as > well as LUCENE_XY ... Starting with 4.9, it also understands LUCENE_X_Y > ... which was a required addition because version 4.10 wouldn't have > worked with the LUCENE_XY notation. > > Thanks, > Shawn > > -- Vincenzo D'Amore email: v.dam...@gmail.com skype: free.dev mobile: +39 349 8513251
Re: "More" facets or "Less" facets in the Velocity UI
This is strictly a front end "development" . You need to modify the velocity template to provide that feature UI side . Solr side you configure your facet.limit ( on a per field basis if necessary). In the "browse" example it's in the appended params under the specific request handler. Cheers 2015-06-23 15:53 GMT+01:00 Sznajder ForMailingList : > Hi > > Is there a way to add a link "More" (or "Less") under the list of facet's > values, in aim to display more facets values? > > Thanks! > > Benjamin > -- -- Benedetti Alessandro Visiting card : http://about.me/alessandro_benedetti "Tyger, tyger burning bright In the forests of the night, What immortal hand or eye Could frame thy fearful symmetry?" William Blake - Songs of Experience -1794 England
Re: Implicit Router Configurations
Please raise a JIRA for this, I can see why this would occur. You can manually change the clusterstate.json file when this happens as a stop-gap, I'd have all the Solr instances down when doing this though. Best, Erick On Tue, Jun 23, 2015 at 8:19 AM, Arnon Yogev wrote: > We have a use case where documents are indexed in shards according to a > specific field (shard per user), and the number of shards is unknown when > creating the collection. > For that purpose we use the implicit router and define router.field=user. > > From what we've seen, the only way to define an implicit router is during > the collection creation. > Moreover, the router definitions (router.name and router.field) are kept > only in clusterstate.json and not in any solr configuration file on disk. > > In some cases solr state becomes inconsistent and we need to delete the > configs from ZK and restart the solr server. The behavior we see is the > new clusterstate.json generated by solr on startup has the default > router.name=compositeId, which is not what we defined during creation. > > Are we missing something? Is there a place to configure the implicit > router on disk such that it will be persistent? > > Thanks, > Arnon > >
Creating A User Interface On Top of Solr
Hello, I'm trying to custom build my own Solr interface in Visual Studios instead of using/modifying the original Velocity interface. I'm mostly doing this as a learning exercise for building UI that's why I'm opting out of using it. The problem is I'm pretty new and not sure where to begin. Most of the posts I've seen end in people opting to modify the default UI. Are there any good links to tutorials or good Solr classes I should look at as I'm trying to build this UI? This interface isn't to replace the Solr Admin page, it's simply to handle the querying of the indexed documents that I have. If this seems more an "exercise in futility" (due to vast amounts of code that needs written or whatnot) then please let me know. I don't really have any reason to believe it should be but I've been wrong before. Thanks in advance -- View this message in context: http://lucene.472066.n3.nabble.com/Creating-A-User-Interface-On-Top-of-Solr-tp4213541.html Sent from the Solr - User mailing list archive at Nabble.com.
Re: Creating A User Interface On Top of Solr
First, the Velocity UI was never intended to be a user-facing application, if for no other reason than it has direct access to Solr. And I can delete all the docs in the collection, delete collections, create new collections and all manner of other really bad stuff. So almost _every_ application has its own UI for search. You don't need to worry in the least about the Solr code base. You shouldn't have to care. All you need to do is send HTTP requests to Solr and deal with the response. For instance, try these two commands from a browser: http://sever:port/solr/collection/query?q=*:* (that's star colon star in case your e-mail makes the colon just a bold). http://sever:port/solr/collection/select?q=*:* You should see a JSON and XML response packet. _That's_ all you really have to deal with. The rest of theUI has nothing to do with Solr and everything to do with your UI skills, you "merely" have to parse the response and display it pleasingly. If you don't want to curl the commands or use your own HTTP client, you can use SolrJ to query Solr and not have to parse the response because there are a bunch of helper methods that make navigating the response easier, google "solrj query example" and you'll see a bunch of examples. Be a little careful to make sure the example you use is compatible with your Solr version. HTH, Erick On Tue, Jun 23, 2015 at 5:20 PM, Paden wrote: > Hello, > > I'm trying to custom build my own Solr interface in Visual Studios instead > of using/modifying the original Velocity interface. I'm mostly doing this as > a learning exercise for building UI that's why I'm opting out of using it. > > The problem is I'm pretty new and not sure where to begin. Most of the posts > I've seen end in people opting to modify the default UI. Are there any good > links to tutorials or good Solr classes I should look at as I'm trying to > build this UI? > > This interface isn't to replace the Solr Admin page, it's simply to handle > the querying of the indexed documents that I have. > > If this seems more an "exercise in futility" (due to vast amounts of code > that needs written or whatnot) then please let me know. I don't really have > any reason to believe it should be but I've been wrong before. > > Thanks in advance > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/Creating-A-User-Interface-On-Top-of-Solr-tp4213541.html > Sent from the Solr - User mailing list archive at Nabble.com.
Re: Same query, inconsistent result in SolrCloud
Dear Erick,
Thank you, I fond it's the problem of my text segmentation setting.
Anyway, thanks.
Regards,
Jerome
2015-06-21 0:43 GMT+08:00 Erick Erickson :
> Just that this _shouldn't_ be going on at all. Either
> 1> you've done something when setting up this collection that's
> producing this incorrect result
> or
> 2> you've found something really bad.
>
> So:
> 1> How did you create your collection? Did you use the Collections API
> or try to define individual cores with the Core Admin api? If the latter,
> you likely missed some innocent-seeming property on the core and your
> collection isn't correctly configured. Please do NOT use the "core admin
> API"
> in SolrCloud unless you know the code very, very well. Use the
> Collections API always.
>
> 2> Try querying each replica with &distrib=false. That'll return only the
> docs on the particular replica you fire the query at. Do you have
> replicas in the _same_ shard with different numbers of docs? If so,
> what can you see that's different about those cores?
>
> 3> What does your clusterstate.json file show?
>
> 4> how did you index documents?
>
> Best,
> Erick
>
> On Fri, Jun 19, 2015 at 8:07 PM, Jerome Yang
> wrote:
> > Hi!
> >
> > I'm facing a problem.
> > I'm using SolrCloud 4.10.3, with 2 shards, each shard have 2 replicas.
> >
> > After index data to the collection, and run the same query,
> >
> > http://localhost:8983/solr/catalog/select?q=a&wt=json&indent=true
> >
> > Sometimes, it return the right,
> >
> > {
> > "responseHeader":{
> > "status":0,
> > "QTime":19,
> > "params":{
> > "indent":"true",
> > "q":"a",
> > "wt":"json"}},
> > "response":{"numFound":5,"start":0,"maxScore":0.43969032,"docs":[
> > {},{},...
> >
> > ]
> >
> > }
> >
> > }
> >
> > But, when I re-run the same query, it return :
> >
> > {
> > "responseHeader":{
> > "status":0,
> > "QTime":14,
> > "params":{
> > "indent":"true",
> > "q":"a",
> > "wt":"json"}},
> > "response":{"numFound":0,"start":0,"maxScore":0.0,"docs":[]
> > },
> > "highlighting":{}}
> >
> >
> > Just some short word will show this kind of problem.
> >
> > Do anyone know what's going on?
> >
> > Thanks
> >
> > Regards,
> >
> > Jerome
>
Re: Implicit Router Configurations
Thank you Erick, What is the recommended way to manually change clusterstate.json? Is there a java code \ script way of editing a file in ZK? Best, Arnon From: Erick Erickson To: solr-user@lucene.apache.org Date: 23/06/2015 09:09 PM Subject:Re: Implicit Router Configurations Please raise a JIRA for this, I can see why this would occur. You can manually change the clusterstate.json file when this happens as a stop-gap, I'd have all the Solr instances down when doing this though. Best, Erick On Tue, Jun 23, 2015 at 8:19 AM, Arnon Yogev wrote: > We have a use case where documents are indexed in shards according to a > specific field (shard per user), and the number of shards is unknown when > creating the collection. > For that purpose we use the implicit router and define router.field=user. > > From what we've seen, the only way to define an implicit router is during > the collection creation. > Moreover, the router definitions (router.name and router.field) are kept > only in clusterstate.json and not in any solr configuration file on disk. > > In some cases solr state becomes inconsistent and we need to delete the > configs from ZK and restart the solr server. The behavior we see is the > new clusterstate.json generated by solr on startup has the default > router.name=compositeId, which is not what we defined during creation. > > Are we missing something? Is there a place to configure the implicit > router on disk such that it will be persistent? > > Thanks, > Arnon > >
