Skip to content

RDoc-4087 Corax: CacheSizeInMb and its behavior on encrypted databases #2690

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Danielle9897 wants to merge 2 commits into
base: main
Choose a base branch
from
Open

RDoc-4087 Corax: CacheSizeInMb and its behavior on encrypted databases #2690

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
81d89e9
RDoc-4087 Corax: CacheSizeInMb and its behavior on encrypted databases
Danielle9897 Oct 2, 2026
ed66f99
RDoc-4087 fix review comments
Danielle9897 Oct 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 13 additions & 0 deletions docs/ai-integration/vector-search/overview.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,6 +31,7 @@ import Panel from '@site/src/components/Panel';
* [Vector search in RavenDB](../../ai-integration/vector-search/overview.mdx#vector-search-in-ravendb)
* [Vector search using a dynamic query](../../ai-integration/vector-search/overview.mdx#vector-search-using-a-dynamic-query)
* [Vector search using a static-index](../../ai-integration/vector-search/overview.mdx#vector-search-using-a-static-index)
* [Vector search on encrypted databases](../../ai-integration/vector-search/overview.mdx#vector-search-on-encrypted-databases)

</Admonition>

Expand Down Expand Up @@ -246,5 +247,17 @@ It explains how to define the index, configure indexing & query-time parameters,
</Admonition>

</Admonition>

<Admonition type="note" title="">

## Vector search on encrypted databases

* Vector queries on [encrypted databases](../../server/security/encryption/database-encryption.mdx#vector-search-node-cache) run without the HNSW node cache.
Without the cache, vector queries may require more disk reads, which can reduce query performance.

* Disabling the cache does not change query results.
See [Indexing.Corax.VectorSearch.CacheSizeInMb](../../server/configuration/indexing-configuration.mdx#indexingcoraxvectorsearchcachesizeinmb) for cache configuration.

</Admonition>

</Panel>
1 change: 0 additions & 1 deletion docs/indexes/index-administration.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,7 +78,6 @@ import LanguageContent from "@site/src/components/LanguageContent";

##### Reset index
* Resetting an index will force re-indexing of all documents that match the index definition.
An index usually needs to be reset once it reached its error quota and is in an _Error_ state.
An index usually needs to be reset when it reaches its error quota and enters the _Error_ state.

* See [reset index](../client-api/operations/maintenance/indexes/reset-index.mdx) for detailed information.
Expand Down
22 changes: 22 additions & 0 deletions docs/server/configuration/indexing-configuration.mdx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,6 +68,7 @@ import LanguageContent from "@site/src/components/LanguageContent";
[Indexing.Corax.MaxMemoizationSizeInMb](../../server/configuration/indexing-configuration.mdx#indexingcoraxmaxmemoizationsizeinmb)
[Indexing.Corax.Static.ComplexFieldIndexingBehavior](../../server/configuration/indexing-configuration.mdx#indexingcoraxstaticcomplexfieldindexingbehavior)
[Indexing.Corax.UnmanagedAllocationsBatchSizeLimitInMb](../../server/configuration/indexing-configuration.mdx#indexingcoraxunmanagedallocationsbatchsizelimitinmb)
[Indexing.Corax.VectorSearch.CacheSizeInMb](../../server/configuration/indexing-configuration.mdx#indexingcoraxvectorsearchcachesizeinmb)
[Indexing.Corax.VectorSearch.DefaultMinimumSimilarity](../../server/configuration/indexing-configuration.mdx#indexingcoraxvectorsearchdefaultminimumsimilarity)
[Indexing.Corax.VectorSearch.DefaultNumberOfCandidatesForIndexing](../../server/configuration/indexing-configuration.mdx#indexingcoraxvectorsearchdefaultnumberofcandidatesforindexing)
[Indexing.Corax.VectorSearch.DefaultNumberOfCandidatesForQuerying](../../server/configuration/indexing-configuration.mdx#indexingcoraxvectorsearchdefaultnumberofcandidatesforquerying)
Expand Down Expand Up @@ -603,6 +604,27 @@ The maximum amount of memory in megabytes that Corax can use for memoization dur



## Indexing.Corax.VectorSearch.CacheSizeInMb

* The memory budget (in MB) used to size the HNSW node cache for each vector field in an index.
The cache keeps the upper-level nodes of the HNSW graph in memory, so vector queries read less from disk.

* Set to `0` to disable the cache.
Changing this value does not require re-indexing.

* The cache is not supported on [encrypted databases](../../server/security/encryption/database-encryption.mdx#vector-search-node-cache).
This setting is ignored on encrypted databases, and vector search reads the HNSW graph from disk.

* This setting is available from RavenDB 7.2.5.

---

- **Type**: `int`
- **Default**: `10`
- **Scope**: Server-wide, or per database, or per index



## Indexing.Corax.VectorSearch.DefaultMinimumSimilarity

* The default minimum similarity to use when making a vector search query.
Expand Down
Loading
Loading