Patrice Bourgougnon
Patrice Bourgougnon
I've had the exact same issue. Thanks to @carlobeltrame , I quickly found the cause. Else, I could have spent a long time, or even give up on this great...
- It happens with 2 sort fields only. - Class has default docker settings (one shard I guess). - Version was 'latest' at the time of the issue. - id...
wpsolr__price_f is set with values (float arrays), and aggregates are fine except when a filter and min/max are both used.
It looks good. Same query now gives (which shows that "stats_filters_array_float_ko" is now fine): ``` { "data": { "search": { "Text2vectransformers1": [ { "wpsolr__price_f": [ 20 ], "wpsolr_pid": 46, "wpsolr_title":...
Is there something new on the admin APIs of Solr ?
How can I see the whole truncated error message from Vespa? ```(Vespa) "Client error: `POST http://node1.vespanet:8080/document/v1/Dev8083B5C559E9F8Dev8083B5C559E9F8/Dev8083B5C559E9F8Dev8083B5C559E9F8/docid/13?create=true` resulted in a `400 Bad Request` response: {"pathId":"/document/v1/Dev8083B5C559E9F8Dev8083B5C559E9F8/Dev8083B5C559E9F8Dev8083B5C559E9F8/docid/13","message":"No fi (truncated...) "```
@kkraune Thanks for your help. It looks like a found a workaround: I detect the error and retry. The "no fi (truncated" is probably "no fields": I create a schema...
I renamed the issue.
You can index a missing field, you should get the same error. The fact that my schema name was long, as it is included in the message, made the truncation...