4509: Rest embedder r=ManyTheFish a=dureuill
Fixes#4531
See [Usage page](https://meilisearch.notion.site/v1-8-AI-search-API-usage-135552d6e85a4a52bc7109be82aeca42?pvs=25#e6f58c3b742c4effb4ddc625ce12ee16)
### Implementation changes
- Remove tokio, futures, reqwests
- Add a new `milli::vector::rest::Embedder` embedder
- Update OpenAI and Ollama embedders to use the REST embedder internally
- Make Embedder::embed a sync method
- Add the new embedder source as described in the usage
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
4530: fix: set the histogram bucket boundaries to follow the otel spec r=curquiza a=rohankmr414
# Pull Request
## What does this PR do?
- Fixes the http request duration histogram bucket boundaries to follow the opentelemetry spec, currently the bucket boundaries are too granular and only track latencies below 1s.
## PR checklist
Please check if your PR fulfills the following requirements:
- [ ] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [x] Have you read the contributing guidelines?
- [x] Have you made sure that the title is accurate and descriptive of the changes?
Thank you so much for contributing to Meilisearch!
Co-authored-by: Rohan Kumar <rohankmr414@gmail.com>
4476: Make the `/facet-search` route use the `sortFacetValuesBy` setting r=irevoire a=Kerollmops
This PR fixes#4423 by ensuring that the `/facet-search` route uses the `sortFacetValuesBy` setting.
Note for the documentation team (to be moved in the tracking issue): Using the new `sortFacetValuesBy` setting can slow down the facet-search requests as Meilisearch iterates over the whole list of facet values and computes the count of documents on every entry. That is hardly or even impossible to optimize correctly.
### TODO
- [x] Create a custom HashMap wrapper for the facet `OrderBy` settings.
This wrapper will return the `OrderBy` setting of the facet, if not defined will use the default `*` one, and if not there either (strange) will fall back on the lexicographic one.
- [x] Create a `ValuesCollection` wrapper that implements the logic for the lexicographic and count order by.
- [x] Use it when there is no search query.
- [x] Use it when there is a search query with and without allowed typos.
- [x] Do not change the original logic, only use a wrapper.
- [x] Add tests
Co-authored-by: Clément Renault <clement@meilisearch.com>
4456: Add Ollama as an embeddings provider r=dureuill a=jakobklemm
# Pull Request
## Related issue
[Related Discord Thread](https://discord.com/channels/1006923006964154428/1211977150316683305)
## What does this PR do?
- Adds Ollama as a provider of Embeddings besides HuggingFace and OpenAI under the name `ollama`
- Adds the environment variable `MEILI_OLLAMA_URL` to set the embeddings URL of an Ollama instance with a default value of `http://localhost:11434/api/embeddings` if no variable is set
- Changes some of the structs and functions in `openai.rs` to be public so that they can be shared.
- Added more error variants for Ollama specific errors
- It uses the model `nomic-embed-text` as default, but any string value is allowed, however it won't automatically check if the model actually exists or is an embedding model
Tested against Ollama version `v0.1.27` and the `nomic-embed-text` model.
## PR checklist
Please check if your PR fulfills the following requirements:
- [x] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [x] Have you read the contributing guidelines?
- [x] Have you made sure that the title is accurate and descriptive of the changes?
Co-authored-by: Jakob Klemm <jakob@jeykey.net>
Co-authored-by: Louis Dureuil <louis.dureuil@gmail.com>
4445: Add subcommand to run benchmarks r=irevoire a=dureuill
# Pull Request
## Related issue
Not user-facing, no issue
## What does this PR do?
- Adds a new `cargo xtask bench` subcommand that can run one or multiple workload files and report the results to a server
- A workload file is a JSON file with a specific schema
- Refactor our use of the `vergen` crate:
- update to the beta `vergen-git2` crate
- VERGEN_GIT_SEMVER_LIGHTWEIGHT => VERGEN_GIT_DESCRIBE
- factor logic in a single `build-info` crate that is used both by meilisearch and xtask (prevents vergen variables from overriding themselves)
- checked that defining the variables by hand when no git repo is available (docker build case) still works.
- Add CI to run `cargo xtask bench`
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
Initial prototype of Ollama embeddings actually working, error handlign / retries still missing.
Allow model to be any String and require dimensions parameter
Fixed rustfmt formatting issues
There were some formatting issues in the initial PR and this should not make the changes comply with the Rust style guidelines
Because I accidentally didn't follow the style guide for commits in my commit messages I squashed them into one to comply
4433: Enhance facet incremental r=Kerollmops a=ManyTheFish
# Pull Request
## Related issue
Fixes#4367Fixes#4409
## What does this PR do?
- Add a test reproducing #4409
- Fix#4409 by removing a document from a level only if it is no more present in all the linked sub-level nodes
- Optimize facet Incremental indexing by creating or deleting a complete level once per field id instead of for each facet value
- Optimize facet Incremental indexing by doing the additions and the deletions in the same process instead of doing them separately
Co-authored-by: ManyTheFish <many@meilisearch.com>
4443: Add GPU analytics r=dureuill a=dureuill
# Pull Request
## Related issue
Adds analytics indicating whether Meilisearch was compiled with the `milli/cuda` feature.
Cc `@macraig`
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
4042: Implements the new replication parameters r=ManyTheFish a=irevoire
### This PR implements the necessary parameters for the High Availability
- [ ] Update the spec
Introduce a new CLI flag called `--experimental-replication-parameters` that changes a few behaviors in the engine:
- [The auto-deletion of tasks is disabled](https://specs.meilisearch.com/specifications/text/0060-tasks-api.html#_2-technical-details)
- Upon registering a task, you can choose its task ID by sending a new header: `TaskId: 456645`. It must be a valid number, which must be superior to the last task id ever seen.
- Add the ability to « dry-register » a task. That means meilisearch will answer to you with a valid task ID like everything went well, but won’t actually write anything in the database. To do that, you need to use the `DryRun: true` header.
----
Old prototype `prototype-custom-task-id-0`:
- Adds the capability to specify your own task ID via the `TaskId` http header
- Make the task IDs a u64 instead of a u32
Co-authored-by: Tamo <tamo@meilisearch.com>
4410: Implement the experimental log mode cli flag and log level updates at runtime r=dureuill a=irevoire
# Pull Request
This PR fixes two issues at once because they’re highly correlated in the codebase.
## Related issue
Fixes https://github.com/meilisearch/meilisearch/issues/4415
Fixes https://github.com/meilisearch/meilisearch/issues/4413
## What does this PR do?
- It makes the fmt logger configurable to output json or human-readable logs (like we already do today)
- It moves the fmt logger under a `reload` layer so we can update its targets at runtime
- Add the possibility to stream logs in the json mode
- Adds an analytics for the new CLI flag
Co-authored-by: Tamo <tamo@meilisearch.com>
4313: Fix document formatting performances r=Kerollmops a=ManyTheFish
reduce the formatted option list to the attributes that should be formatted,
instead of all the attributes to display.
The time to compute the `format` list scales with the number of fields to format;
cumulated with `map_leaf_values` that iterates over all the nested fields, it gives a quadratic complexity:
`d*f` where `d` is the total number of fields to display and `f` is the total number of fields to format.
Co-authored-by: ManyTheFish <many@meilisearch.com>
4279: Check experimental feature on setting update query rather than in the task. r=ManyTheFish a=dureuill
Improve the UX by checking for the vector store feature and returning an error synchronously when sending a setting update, rather than in the indexing task.
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
4238: Task queue webhook r=dureuill a=irevoire
# Prototype `prototype-task-queue-webhook-1`
The prototype is available through Docker by using the following command:
```bash
docker run -p 7700:7700 -v $(pwd)/meili_data:/meili_data getmeili/meilisearch:prototype-task-queue-webhook-1
```
# Pull Request
Implements the task queue webhook.
## Related issue
Fixes https://github.com/meilisearch/meilisearch/issues/4236
## What does this PR do?
- Provide a new cli and env var for the webhook, respectively called `--task-webhook-url` and `MEILI_TASK_WEBHOOK_URL`
- Also supports sending the requests with a custom `Authorization` header by specifying the optional `--task-webhook-authorization-header` CLI parameter or `MEILI_TASK_WEBHOOK_AUTHORIZATION_HEADER` env variable.
- Throw an error if the specified URL is invalid
- Every time a batch is processed, send all the finished tasks into the webhook with our public `TaskView` type as a JSON Line GZIPed body.
- Add one test.
## PR checklist
### Before becoming ready to review
- [x] Add a test
- [x] Compress the data we send
- [x] Chunk and stream the data we send
- [x] Remove the unwrap in the index-scheduler when sending the data fails
- [x] The analytics are missing
### Before merging
- [x] Release a prototype
Co-authored-by: Tamo <tamo@meilisearch.com>
Co-authored-by: Clément Renault <clement@meilisearch.com>
4277: Update mini-dashboard to v0.2.12 r=curquiza a=mdubus
# Pull Request
## Related issue
Fixes#4276
## What does this PR do?
Upgrade mini-dashboard to version 0.2.12 ([see changes](https://github.com/meilisearch/mini-dashboard/releases/tag/v0.2.12))
## PR checklist
Please check if your PR fulfills the following requirements:
- [x] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [x] Have you read the contributing guidelines?
- [x] Have you made sure that the title is accurate and descriptive of the changes?
Thank you so much for contributing to Meilisearch!
Co-authored-by: Morgane Dubus <30866152+mdubus@users.noreply.github.com>
- DistributionShift in Search object (to be set from model in embed?)
- Fix issue where embedder index wasn't computed at search time
- Accept as default embedder either the "default" one, or the only embedder when there is only one
4223: Update to heed 0.20 r=dureuill a=Kerollmops
This PR brings the v0.20-alpha.9 version of heed into Meilisearch 🎉 The main goal is to test it in a real environment to make the necessary changes if needed. We also want to merge it as soon as possible during the pre-release phase to ensure we catch bugs before the release.
Most of the calls to heed are the same as before, except:
- The `PolyDatabase` has been replaced with a `Database<Unspecified, Unspecified>`. We replaced the `get<T, U>()` by a `remap<T, U>().get()` calls.
- The `Database` `append(...)` method has been replaced with a `put_with_flags(PutFlags::APPEND, ...)`.
- The `RwTxn<'e, 'p>` has been simplified into a `RwTxn<'e>`.
- The `BytesEncode/Decode` traits return a `Result<_, BoxedError>` instead of an `Option<_>`.
- We no longer need to wrap and unwrap the `BEU32` integer when storing/getting them from heed.
### TODO
- [x] Create actual, simple error types instead of using strings in the codecs.
### Follow-up work
- Move the codecs into another member crate (we depend on the uuid one in the meilitool crate).
- Display the internal decoding error in the `SerializationError` internal error variant.
Co-authored-by: Clément Renault <clement@meilisearch.com>
4126: Make the experimental route /metrics activable via HTTP r=dureuill a=braddotcoffee
# Pull Request
## Related issue
Closes#4086
## What does this PR do?
- [x] Make `/metrics` available via HTTP as described in #4086
- [x] The users can still launch Meilisearch using the `--experimental-enable-metrics` flag.
- [x] If the flag `--experimental-enable-metrics` is activated, a call to the `GET /experimental-features` route right after the launch will show `"metrics": true` even if the user has not called the `PATCH /experimental-features` route yet.
- [x] Even if the --experimental-enable-metrics flag is present at launch, calling the `PATCH /experimental-features` route with `"metrics": false` disables the experimental feature.
- [x] Update the spec
- I was unable to find docs in this repository to update about the `/experimental-features` endpoint. I'll happily update if you point me in the right direction!
## PR checklist
Please check if your PR fulfills the following requirements:
- [x] Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
- [x] Have you read the contributing guidelines?
- [x] Have you made sure that the title is accurate and descriptive of the changes?
Co-authored-by: bwbonanno <bradfordbonanno@gmail.com>
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
The issue was that the operation « DocumentDeletionByFilter » was not
declared as an index operation. That means the indexes stats were not
reprocessed after the application of the operation.
3994: Fix synonyms with separators r=Kerollmops a=ManyTheFish
# Pull Request
## Related issue
Fixes#3977
## Available prototype
```
$ docker pull getmeili/meilisearch:prototype-fix-synonyms-with-separators-0
```
## What does this PR do?
- add a new test
- filter the empty synonyms after normalization
Co-authored-by: ManyTheFish <many@meilisearch.com>
3955: Update mini-dashboard to version 0.2.11 r=curquiza a=bidoubiwa
# Pull Request
## What does this PR do?
- Updates the mini-dashboard to version [0.2.11](https://github.com/meilisearch/mini-dashboard/releases/tag/v0.2.11)
## PR checklist
- [x] Have you read the contributing guidelines?
- [x] Have you made sure that the title is accurate and descriptive of the changes?
Thank you so much for contributing to Meilisearch!
Co-authored-by: Charlotte Vermandel <charlottevermandel@gmail.com>
3953: Update UTM campaign r=curquiza a=macraig
# Pull Request
## What does this PR do?
Redirect CTAs to Cloud landing page
Co-authored-by: María <maria@Marias-MacBook-Pro.local>
3942: Normalize for the search the facets values r=ManyTheFish a=Kerollmops
This PR improves and fixes the search for facet values feature. Searching for _bre_ wasn't returning facet values like _brévent_ or _brô_.
The issue was related to the fact that facets are normalized but not in the same way as the `searchableAttributes` are. We decided to normalize them further and add another intermediate database where the key is the normalized facet value, and the value is a set of the non-normalized facets. We then use these non-normalized ones to get the correct counts by fetching the associated databases.
### What's missing in this PR?
- [x] Apply the change to the whole set of `SearchForFacetValue::execute` conditions.
- [x] Factorize the code that does an intermediate normalized value fetch in a function.
- [x] Add or modify the search for facet value test.
Co-authored-by: Clément Renault <clement@meilisearch.com>
Co-authored-by: Kerollmops <clement@meilisearch.com>
3933: Stop computing the update files size r=ManyTheFish a=Kerollmops
This PR, related #3934, removes the part which computes the total size of the `data.ms/update_files` folder, which can take a lot of time when many updates must be processed.
It is not breaking API-side but is breaking on the result we will show to the user. The `databaseSize` field returned by the `/stats` endpoint will be reduced.
Co-authored-by: Kerollmops <clement@meilisearch.com>
3921: Deactivate camel case segmentation r=dureuill a=ManyTheFish
# Pull Request
This PR deactivates the camel case segmentation to retrieve the possibility to accept typos over camel-cased words
## Related issue
Fixes#3869Fixes#3818
## What does this PR do?
- deactivates camelcase segmentation
related to #3919
Co-authored-by: ManyTheFish <many@meilisearch.com>
3907: Add telemetry for define field to search on at query time r=dureuill a=ManyTheFish
Add "attributes_to_search_on" telemetry usage counter:
```json
"attributes_to_search_on": {
"total_number_of_use": 12,
},
```
This measures the number of search queries that the user uses `attributesToSearchOn` field.
related to https://github.com/meilisearch/specifications/pull/251
## reviewers:
- `@macraig` for validating the telemetry's name
- `@dureuill` for validating the code
Co-authored-by: ManyTheFish <many@meilisearch.com>
3897: Add automated tests for `/experimental-features` route r=Kerollmops a=dureuill
# Pull Request
## What does this PR do?
- Make `RuntimeTogglableFeatures` `Eq`
- Add various tests for the `/experimental-features` route
- Integration tests for the route itself
- Integration tests for the effect of enabling `scoreDetails` and `vectorStore` through this route.
- Dump integration tests
Co-authored-by: Louis Dureuil <louis@meilisearch.com>
3889: Display the total number of tasks matching a filter/query r=dureuill a=Kerollmops
This PR returns a new field on the `/tasks` routes. The `total` field exposes the total number of tasks that matches the given filter/query. It is useful to display information on a user interface and can help understand when progress is made in processing tasks, i.e., the total number of tasks on `/tasks?statuses=succeeded` will increase over time.
Fixes#3888.
- [ ] Update the specs fo the `/tasks` route.
## How have I implemented it?
I found it much easier to run two times the task filtering system. Once with the original `from` and `limit` parameters and a second time without. The second call will return the total number of tasks that match the query, not only the number of tasks on the current page.
So far, in terms of performance, there doesn't seem to be any issue. I tried different filters with something like 250k tasks. Note that there is a limit of 1M tasks in the queue.
Co-authored-by: Clément Renault <clement@meilisearch.com>
3891: Fix the way we compute the 99th percentile r=dureuill a=Kerollmops
This PR fixes how we compute the 99th percentile by avoiding using float and doing the multiplication and divisions in the correct order avoiding going out of the buffer of timings. You can see the issue on [this rust playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2021).
When there are a very small number of successful requests, the number is so tiny that the 99th percentile calculus sometimes gives an index out of the buffer. In this example, the `1`/`1.0` represent the number of timings you collected (one). As you can see, the float computation gives us the index `1.0`, with is out of a vector of only one value. This makes the engine generate a `null` value.
```rust
1 * 99 / 100 = 0 // with integers
0.99_f64 * (1.0 - 1.0) + 1.0 = 1.0 // with floats
```
Co-authored-by: Clément Renault <clement@meilisearch.com>
