[
https://issues.apache.org/jira/browse/SOLR-17950?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ishan Chattopadhyaya updated SOLR-17950:
----------------------------------------
Summary: Explore using FlatVectorsWriter for DenseVectorsField's
stored=true (was: Explore using FlatVectorsWriter for DenseVectorsWriter's
stored=true)
> Explore using FlatVectorsWriter for DenseVectorsField's stored=true
> -------------------------------------------------------------------
>
> Key: SOLR-17950
> URL: https://issues.apache.org/jira/browse/SOLR-17950
> Project: Solr
> Issue Type: Improvement
> Reporter: Ishan Chattopadhyaya
> Priority: Major
>
> Today, when indexed=true, all vectors are written to a FlatVectorsWriter and
> a HNSW graph is built using that. When stored=true, a regular stored field is
> built using that. For a large vector index, it is wasteful to store raw
> vectors in two places. Can we retrieve the vectors from the flat vectors
> index instead of the stored fields to save on disk space and indexing time?
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
