For in-place updates you should read this: https://lucene.apache.org/solr/guide/8_1/updating-parts-of-documents.html
On Mon, May 11, 2020 at 2:49 PM Vincenzo D'Amore <v.dam...@gmail.com> wrote: > But keep in mind that "With the exception of in-place updates, the whole > block must be updated or deleted together, not separately. For some > applications this may result in tons of extra indexing and thus may be a > deal-breaker." > > On Mon, May 11, 2020 at 1:37 PM Jan Høydahl <jan....@cominvent.com> wrote: > >> Sounds like you are looking for parent/child docs here, see >> https://lucene.apache.org/solr/guide/8_5/indexing-nested-documents.html >> >> { >> "type": "user", >> "name": "user1", >> "products": [ >> { "id": "prod_A", "cost": 50}, >> { "id": "prod_B", "cost": 200}, >> { "id": "prod_D", "cost": 25} >> ] >> } >> >> This will index 4 documents - one user document and three product-cost >> child documents. >> >> You can then search the child docs and return matching parents with e.g. >> q=*:*&fq={!parent which="type:user"}((id:prod_A AND cost:[50 TO 100]) OR >> (id:prod_D AND cost:[0 TO 40]))&fl=[child] >> >> Hope this helps. >> >> Jan >> >> > 11. mai 2020 kl. 11:35 skrev Vignan Malyala <dsmsvig...@gmail.com>: >> > >> > I have around 1M products used by my clients. >> > Client need a filter of these 1M products by their cost filters. >> > >> > Just like: >> > User1 has 5 products (A,B,C,D,E) >> > User2 has 3 products (D,E,F) >> > User3 has 10 products (A,B,C,H,I,J,K,L,M,N,O) >> > >> > ...every customer has different sets. >> > >> > Now they want to search users by filter of product costs: >> > Product_A_cost : 50 TO 100 >> > Product_D_cost : 0 TO 40 >> > >> > it should return all the users who use products in this filter range. >> > >> > As I have 1M products, do I need to create dynamic fields for all users >> > with filed names as Product_A_cost and product_B_cost..... etc to make a >> > search by them? If I should, then I haveto create 1M dynamic fields >> > Or is there any other way? >> > >> > Hope I'm clear here! >> > >> > >> > On Mon, May 11, 2020 at 1:47 PM Jan Høydahl <jan....@cominvent.com> >> wrote: >> > >> >> Sounds like an anti pattern. Can you explain what search problem you >> are >> >> trying to solve with this many unique fields? >> >> >> >> Jan Høydahl >> >> >> >>> 11. mai 2020 kl. 07:51 skrev Vignan Malyala <dsmsvig...@gmail.com>: >> >>> >> >>> Hi >> >>> Is it good idea to create 100000 dynamic fields of time pint in solr? >> >>> I have that many fields to search on actually which come upon based on >> >>> users. >> >>> >> >>> Thanks in advance! >> >>> And I'm using Solr Cloud in real-time. >> >>> >> >>> Regards, >> >>> Sai Vignan M >> >> >> >> > > -- > Vincenzo D'Amore > > -- Vincenzo D'Amore
