Commit Graph

373 Commits

Author SHA1 Message Date
95821d0bde refactor: update macro 2025-05-26 10:07:13 +02:00
f690fa0686 feat: add macro_rules to factorize 2025-05-26 09:46:14 +02:00
24e94b28c1 feat: uncouple geo extraction from full doc 2025-05-26 09:22:20 +02:00
1d5265caf4 Fix typo in method name 2025-05-22 14:25:04 +00:00
73e4206b3c Pass a progress callback to recompute_word_fst_from_word_docids_database
fixes https://github.com/meilisearch/meilisearch/pull/5494#discussion_r2069377991
2025-05-21 10:49:43 +02:00
a964251cee Remove useless reset
fixes https://github.com/meilisearch/meilisearch/pull/5494#discussion_r2069373494
2025-05-21 10:49:43 +02:00
79db2e67fb refactor: prefer helper over explicit pool construction
Co-authored-by: Many the fish <many@meilisearch.com>
2025-05-15 11:24:34 +02:00
865f24cfef refactor: helper methods for pool and max threads 2025-05-14 23:45:24 +02:00
15cdc6924b refactor: remove runtime cfg!(test) check
Won't work in integration tests and consequently all threads would be
used. To remedy this we make explicit `max_threads=Some(1)` in the
IndexerConfig::default
2025-05-13 09:18:19 +02:00
75a7e40a27 Merge branch 'main' into all-cpus-in-import-dump 2025-05-12 21:48:12 +02:00
d99419acfb Add a NoOp operation in index update 2025-05-12 14:19:15 +02:00
53f32a7dd7 refactor: change thread_pool from Option<ThreadPoolNoAbort> to
ThreadPoolNoAbort
2025-05-07 17:00:08 +02:00
47a7ed93d3 feat: Make MaxThreads None by default 2025-05-06 09:11:55 +02:00
71ab11f1fe Merge pull request #5523 from meilisearch/rollback-updates
Allow rollbacking updates
2025-05-05 09:53:56 +00:00
2ac826edca Apply suggested changes
Co-authored-by: Clément Renault <renault.cle@gmail.com>

Update crates/meilisearch/src/lib.rs

Co-authored-by: Clément Renault <renault.cle@gmail.com>
2025-05-01 16:12:06 +02:00
8b23eddc10 Dumpless upgrade 2025-04-30 18:03:50 +02:00
89aff2081c Fix clippy warnings 2025-04-30 14:17:32 +02:00
0cb2bf34a5 Fix test 2025-04-29 14:47:30 +02:00
3b773b3416 Revert thread_pool type back to Option in config 2025-04-28 11:56:37 +02:00
648b2876f6 Create temp threadpool with all CPUs in dump 2025-04-27 00:52:10 +02:00
121c1ac1dd Upgrade supports cancelling 2025-04-24 16:08:10 +02:00
49add50cb3 Make version constants u32 2025-04-24 16:06:19 +02:00
63a4dfa2a8 Add disableOnNumber setting 2025-04-23 16:57:50 +02:00
b025f1bcf1 Merge branch 'main' into release-v1.14.0-tmp 2025-04-14 12:35:47 +02:00
a0bfcf8872 Make cargo fmt happy 2025-04-01 11:27:41 +02:00
64477aac60 Box the large GeoError error variant 2025-04-01 11:26:34 +02:00
4d90e3d2ec Make Cargo and Clippy happy 2025-04-01 11:26:34 +02:00
f729864466 Check dimension mismatch at insertion time 2025-03-31 15:27:49 +02:00
bb2e9419d3 Merge pull request #5468 from meilisearch/more-precise-post-processing
More Precise Post Processing
2025-03-27 10:07:09 +00:00
811143cbe9 Add more progress precision when doing post processing 2025-03-27 10:17:28 +01:00
db7ce03763 Improve the performances of computing the size of the documents database 2025-03-26 17:40:12 +01:00
bf3a29b60d Document problematic case in test and acknowledge PR comment 2025-03-26 12:57:25 +01:00
43c8a206b4 detail comments 2025-03-25 13:07:17 +01:00
6b1c262b74 fix all tests 2025-03-25 12:43:15 +01:00
d71c6f3483 allow multiple embedding in per document per embedder to pass 2025-03-25 12:04:25 +01:00
a09d08c7b6 Avoid reindexing searchable order changes
Update settings.rs

Update settings.rs
2025-03-24 16:26:52 +01:00
e019ad7692 Display more detailed error message instead of panic 2025-03-21 15:41:31 +01:00
cbdf80893d Merge #5422
5422: Add more progress levels to measure merging r=Kerollmops a=Kerollmops

I found out that Meilisearch was not correctly reporting the long indexing times in the progress and that a lot of time was spent on extracting words with all documents already extracted. The reason was that there was no step to report merging the cache and sending the entries to write to the writer thread. This PR adds these entries to the progress.

Co-authored-by: Kerollmops <clement@meilisearch.com>
2025-03-17 12:02:46 +00:00
e2156ddfc7 Simplify the IndexingStep progress enum 2025-03-17 11:40:50 +01:00
13a88d6131 Merge #5407
5407: Geo update bug r=irevoire a=ManyTheFish

# Pull Request

## Related issue
Fixes #5380
Fixes #5399



Co-authored-by: Tamo <tamo@meilisearch.com>
Co-authored-by: ManyTheFish <many@meilisearch.com>
2025-03-17 10:24:33 +00:00
cb16baab18 Add more progress levels to measure merging 2025-03-17 10:13:29 +01:00
009c36a4d0 Add support for the progress API of arroy 2025-03-13 19:00:43 +01:00
e2d372823a Disable the cache by default and make it experimental 2025-03-13 17:22:51 +01:00
5ef7767429 Let arroy uses all the memory available instead of 50% of the 70% 2025-03-13 15:06:03 +01:00
a92a48b9b9 Do not recompute stats on dumpless upgrade
Co-authored-by: Tamo <tamo@meilisearch.com>
2025-03-13 13:58:58 +01:00
d53225bf64 uses a random seed instead of 42 2025-03-13 12:43:31 +01:00
ef9d9f8481 set the memory in arroy 2025-03-13 11:29:00 +01:00
fedb444e66 Fix the upgrade arroy calls 2025-03-13 11:07:49 +01:00
566b4efb06 Dumpless upgrade from v1.13 to v1.14 2025-03-13 11:07:44 +01:00
21bbbdec76 Specify WithoutTls everywhere 2025-03-13 11:07:38 +01:00