No, it doesn't affect facet counts, grouping, or any of that stuff. After
all, facets and grouping are only calculated for documents that
satisfy a query, and deleted documents are, by definition, excluded
from "satisfying a query".
There are some subtle issues in scoring that can be affected, but
Thanks Eric,
This is indeed what we are seeing. I hope we can just ignore the
frequencies. Does it in any way effect facet counts for such records?
Sohail
solr updates are really a delete followed by a re-index. The old terms
are left in the index, but the associated document is marked as deleted.
Schema browser, for instance, will happily report frequencies > 1 for
s when a document has been updated.
You can ignore this if you query on the schemaId
can you post the schema you are applying pls?
On 5 July 2012 11:28, Sohail Aboobaker wrote:
> Another observation is that when we query an individual schemaid, it
> returns only one row using the search interface. Why would frequency be
> more than 1?
Another observation is that when we query an individual schemaid, it
returns only one row using the search interface. Why would frequency be
more than 1?
We have defined the schemaid as String. It has concatenated value of the
product id and language. It takes the form of ID-EN. For example:
'123012-EN', '124020-EN', '12392-FR'.
Sohail
Hello,
Make sure your unique id has a type which always yields one token
after tokenisation is applied (e.g. either "string" or a type which
only defines the KeywordTokenizer in its chain)
Regards,
Savvas
On 5 July 2012 11:02, Sohail Aboobaker wrote:
> Hi,
>
> We have defined a unique key as sc
