Repository navigation
Conversation
Elastic Docs AI PR menuCheck the box to run an AI review for this pull request.
Powered by GitHub Agentic Workflows and docs-actions. For more information, reach out to the docs team. |
🔍 Preview links for changed docs |
✅ Elastic Docs Style Checker (Vale)No issues found on modified lines! The Vale linter checks documentation changes against the Elastic Docs style guide. To use Vale locally or report issues, refer to Elastic style guide for Vale. |
yannis-roussos
left a comment
There was a problem hiding this comment.
Thank you @lcawl, looks great! I have added a few comments and small suggestions
| For every supported value, refer to [`index.mode`](elasticsearch://reference/elasticsearch/index-settings/index-modules.md#index-mode-setting). | ||
|
|
||
| Standard (`standard`) {applies_to}`vectordb: unavailable` | ||
| : Use it for documents you update or delete, and for data that isn't logs, metrics, or vectors. |
There was a problem hiding this comment.
Standard is what Elasticsearch is known for.. you know.. for search :-)
Also, would be better to say "optimized" instead of "use it"? Your call on this :-)
I also removed the cases it's not good for as they are evident from the other specialized modes bellow
| : Use it for documents you update or delete, and for data that isn't logs, metrics, or vectors. | |
| : Optimized for search and any type of retrieval. |
| : Use it for documents you update or delete, and for data that isn't logs, metrics, or vectors. | ||
|
|
||
| Columnar (`columnar`) {applies_to}`stack: preview 9.5` {applies_to}`serverless: preview` {applies_to}`vectordb: unavailable` | ||
| : Use it for analytics-heavy data that isn't logs. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). |
There was a problem hiding this comment.
| : Use it for analytics-heavy data that isn't logs. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). | |
| : Optimized for storage efficiency and analytics. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). |
| : Use it for analytics-heavy data that isn't logs. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). | ||
|
|
||
| LogsDB (`logsdb`) {applies_to}`stack: ga 9.0+` {applies_to}`vectordb: unavailable` | ||
| : Use it to store logs. [Logs data streams](/manage-data/data-store/data-streams/logs-data-stream.md) use this mode. |
There was a problem hiding this comment.
| : Use it to store logs. [Logs data streams](/manage-data/data-store/data-streams/logs-data-stream.md) use this mode. | |
| : Optimized for logs use cases, with log-oriented defaults. [Logs data streams](/manage-data/data-store/data-streams/logs-data-stream.md) use this mode. |
| : Use it to store logs with columnar storage. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). | ||
|
|
||
| TSDB (`time_series`) {applies_to}`vectordb: unavailable` | ||
| : Use it to store metrics. [Time series data streams](/manage-data/data-store/data-streams/time-series-data-stream-tsds.md) use this mode. This mode is for data streams only. |
There was a problem hiding this comment.
| : Use it to store metrics. [Time series data streams](/manage-data/data-store/data-streams/time-series-data-stream-tsds.md) use this mode. This mode is for data streams only. | |
| : Optimized for metrics and time series use cases. [Time series data streams](/manage-data/data-store/data-streams/time-series-data-stream-tsds.md) use this mode. This mode is for data streams only. |
| : Use it to store metrics. [Time series data streams](/manage-data/data-store/data-streams/time-series-data-stream-tsds.md) use this mode. This mode is for data streams only. | ||
|
|
||
| VectorDB (`vectordb_document`) {applies_to}`stack: ga 9.5+` {applies_to}`vectordb: ga` | ||
| : Use it for similarity search on dense vectors. Refer to [Vector index mode](/solutions/search/vector.md#vector-index-mode). In {{vectordb}} projects, it is the only mode available. |
There was a problem hiding this comment.
| : Use it for similarity search on dense vectors. Refer to [Vector index mode](/solutions/search/vector.md#vector-index-mode). In {{vectordb}} projects, it is the only mode available. | |
| : Optimized for vector search use cases, enabling similarity search on dense vectors. Refer to [Vector index mode](/solutions/search/vector.md#vector-index-mode). In {{vectordb}} projects, it is the only mode available. |
| : Use it to store logs. [Logs data streams](/manage-data/data-store/data-streams/logs-data-stream.md) use this mode. | ||
|
|
||
| LogsDB columnar (`logsdb_columnar`) {applies_to}`stack: preview 9.5` {applies_to}`serverless: preview` {applies_to}`vectordb: unavailable` | ||
| : Use it to store logs with columnar storage. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). |
There was a problem hiding this comment.
I am not good with adding correct link paths: We can also add a link to the limitations section (https://www.elastic.co/docs/reference/elasticsearch/columnar#columnar-limitations)
| : Use it to store logs with columnar storage. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). | |
| : Optimized for logs use cases that align with columnar semantics and limitations. Refer to [Columnar index mode](/manage-data/data-store/columnar.md). |
| Standard (`standard`) {applies_to}`vectordb: unavailable` | ||
| : Use it for documents you update or delete, and for data that isn't logs, metrics, or vectors. | ||
|
|
||
| Columnar (`columnar`) {applies_to}`stack: preview 9.5` {applies_to}`serverless: preview` {applies_to}`vectordb: unavailable` |
There was a problem hiding this comment.
I would switch the order slightly and move columnar after TSDB
It feels weird to me to start with columnar as it is not (yet) as important or prominent as other use cases. I would go by order of use for Elasticsearch use cases:
- Standard
- Logs Use cases (first logsdb then logsdb_columnar)
- Metrics / Time series
- then Columnar. I don't think that it matters here that we mention first logsdb columnar and then the pure columnar one. We go by use case
- VectorDB
- Finally and always last: lookup (it's not an equally important index mode, just a wanna-be, an impostor that craves the attention of other modes). Kidding aside, this one does not serve a stand alone use case per se, but acts as a tool, a bridge to enable joining data.
| : Use it for reference data you join in {{esql}}, for example a list of users, hosts, or known threat IPs. Refer to [Join data from multiple indices with LOOKUP JOIN](elasticsearch://reference/query-languages/esql/esql-lookup-join.md). Only the lookup index needs this mode, not the source index. | ||
|
|
||
| Index mode works together with [mappings](/manage-data/data-store/mapping.md) and [templates](/manage-data/data-store/templates.md). | ||
| The mode changes default settings and whether {{es}} stores the original JSON document ([`_source`](/manage-data/data-store/index-basics.md#elasticsearch-intro-documents-fields)) or reconstructs it. |
There was a problem hiding this comment.
(nit) small repetition here of things already mentioned at the start of the section. Maybe it's ok
Summary
The current data-store overview page is a general introduction to indices, mappings, and templates, but does not explain how index modes fit into the higher-level concepts.
Details
The Elasticsearch data store(
manage-data/data-store.md) now has index mode section, placed after the storage overview.Columnar index mode(
manage-data/data-store/columnar.md) and index fundamentals(manage-data/data-store/index-basics.md) now point at that section. Existing LogsDB, TSDS, and "vector index mode" wording elsewhere is unchanged.Generative AI disclosure
Tool(s) and model(s) used: Grok 4.6