It sounds suspiciously similar to https://issues.apache.org/jira/browse/SOLR-3793 which was fixed in Solr 4.0
You should upgrade to a more recent Solr version (4.3.1 is the latest) and see if it's still a problem for you. On Fri, Jun 21, 2013 at 3:19 AM, Samuel García Martínez <samuelgmarti...@gmail.com> wrote: > just to clarify, we send manually the "commit and optimize" we use to fix > this problem. The index process send its own commit, making "searchable" > the new facet values. But it seems that this process is not deleting any > previous value filled used by the uninvertedfield. > > > On Thu, Jun 20, 2013 at 11:42 PM, Samuel García Martínez < > samuelgmarti...@gmail.com> wrote: > >> Hi all, we are getting some facet (faceting a multivalued field) values >> with 0 results using *:* query. I think this is really strange, since we >> are using MatchAllQuery there is no way we can get 0 results in any value. >> >> That 0 results values were present in the index before the reindex we >> made. We fixed it, so far, sending a commit and an optimize. >> >> We still need to obtain facets with 0 results with current values, so >> mincount is not an option. >> >> Solr version: 3.6.0 >> Grouping: false >> post grouping faceting: false >> filter queries: 0 >> >> Is this any known bug or an intended behaviour? Only happens with >> uninvertedfield? >> >> Many thanks :) >> -- >> Un saludo, >> Samuel García. >> > > > > -- > Un saludo, > Samuel García. -- Regards, Shalin Shekhar Mangar.
