609: Retry downloading the benchmarks datasets r=Kerollmops a=irevoire
Downloading the benchmarks datasets is failing [more and more](https://github.com/meilisearch/milli/pull/607#pullrequestreview-1076023074) often; thus, instead of fixing the issue, I thought we could retry multiple times.
Co-authored-by: Irevoire <tamo@meilisearch.com>
596: Filter operators: NOT + IN[..] r=irevoire a=loiclec
# Pull Request
## What does this PR do?
Implements the changes described in https://github.com/meilisearch/meilisearch/issues/2580
It is based on top of #556
Co-authored-by: Loïc Lecrenier <loic@meilisearch.com>
607: Better threshold r=Kerollmops a=irevoire
# Pull Request
## What does this PR do?
Fixes#570
This PR tries to improve the threshold used to trigger the real deletion of documents.
The deletion is now triggered in two cases;
- 10% of the total available space is used by soft deleted documents
- 90% of the total available space is used.
In this context, « total available space » means the `map_size` of lmdb.
And the size used by the soft deleted documents is actually an estimation. We can't determine precisely the size used by one document thus what we do is; take the total space used, divide it by the number of documents + soft deleted documents to estimate the size of one average document. Then multiply the size of one avg document by the number of soft deleted document.
--------
<img width="808" alt="image" src="https://user-images.githubusercontent.com/7032172/185083075-92cf379e-8ae1-4bfc-9ca6-93b54e6ab4e9.png">
Here we can see we have a ~10GB drift in the end between the space used by the soft deleted and the real space used by the documents.
Personally I don’t think that's a big issue because once the red line reach 90GB everything will be freed but now you know.
If you have an idea on how to improve this estimation I would love to hear it.
It look like the difference is linear so maybe we could simply multiply the current estimation by two?
Co-authored-by: Irevoire <tamo@meilisearch.com>
587: Word prefix pair proximity docids indexation refactor r=Kerollmops a=loiclec
# Pull Request
## What does this PR do?
Refactor the code of `WordPrefixPairProximityDocIds` to make it much faster, fix a bug, and add a unit test.
## Why is it faster?
Because we avoid using a sorter to insert the (`word1`, `prefix`, `proximity`) keys and their associated bitmaps, and thus we don't have to sort a potentially very big set of data. I have also added a couple of other optimisations:
1. reusing allocations
2. using a prefix trie instead of an array of prefixes to get all the prefixes of a word
3. inserting directly into the database instead of putting the data in an intermediary grenad when possible. Also avoid checking for pre-existing values in the database when we know for certain that they do not exist.
## What bug was fixed?
When reindexing, the `new_prefix_fst_words` prefixes may look like:
```
["ant", "axo", "bor"]
```
which we group by first letter:
```
[["ant", "axo"], ["bor"]]
```
Later in the code, if we have the word2 "axolotl", we try to find which subarray of prefixes contains its prefixes. This check is done with `word2.starts_with(subarray_prefixes[0])`, but `"axolotl".starts_with("ant")` is false, and thus we wrongly think that there are no prefixes in `new_prefix_fst_words` that are prefixes of `axolotl`.
## StrStrU8Codec
I had to change the encoding of `StrStrU8Codec` to make the second string null-terminated as well. I don't think this should be a problem, but I may have missed some nuances about the impacts of this change.
## Requests when reviewing this PR
I have explained what the code does in the module documentation of `word_pair_proximity_prefix_docids`. It would be nice if someone could read it and give their opinion on whether it is a clear explanation or not.
I also have a couple questions regarding the code itself:
- Should we clean up and factor out the `PrefixTrieNode` code to try and make broader use of it outside this module? For now, the prefixes undergo a few transformations: from FST, to array, to prefix trie. It seems like it could be simplified.
- I wrote a function called `write_into_lmdb_database_without_merging`. (1) Are we okay with such a function existing? (2) Should it be in `grenad_helpers` instead?
## Benchmark Results
We reduce the time it takes to index about 8% in most cases, but it varies between -3% and -20%.
```
group indexing_main_ce90fc62 indexing_word-prefix-pair-proximity-docids-refactor_cbad2023
----- ---------------------- ------------------------------------------------------------
indexing/-geo-delete-facetedNumber-facetedGeo-searchable- 1.00 1893.0±233.03µs ? ?/sec 1.01 1921.2±260.79µs ? ?/sec
indexing/-movies-delete-facetedString-facetedNumber-searchable- 1.05 9.4±3.51ms ? ?/sec 1.00 9.0±2.14ms ? ?/sec
indexing/-movies-delete-facetedString-facetedNumber-searchable-nested- 1.22 18.3±11.42ms ? ?/sec 1.00 15.0±5.79ms ? ?/sec
indexing/-songs-delete-facetedString-facetedNumber-searchable- 1.00 41.4±4.20ms ? ?/sec 1.28 53.0±13.97ms ? ?/sec
indexing/-wiki-delete-searchable- 1.00 285.6±18.12ms ? ?/sec 1.03 293.1±16.09ms ? ?/sec
indexing/Indexing geo_point 1.03 60.8±0.45s ? ?/sec 1.00 58.8±0.68s ? ?/sec
indexing/Indexing movies in three batches 1.14 16.5±0.30s ? ?/sec 1.00 14.5±0.24s ? ?/sec
indexing/Indexing movies with default settings 1.11 13.7±0.07s ? ?/sec 1.00 12.3±0.28s ? ?/sec
indexing/Indexing nested movies with default settings 1.10 10.6±0.11s ? ?/sec 1.00 9.6±0.15s ? ?/sec
indexing/Indexing nested movies without any facets 1.11 9.4±0.15s ? ?/sec 1.00 8.5±0.10s ? ?/sec
indexing/Indexing songs in three batches with default settings 1.18 66.2±0.39s ? ?/sec 1.00 56.0±0.67s ? ?/sec
indexing/Indexing songs with default settings 1.07 58.7±1.26s ? ?/sec 1.00 54.7±1.71s ? ?/sec
indexing/Indexing songs without any facets 1.08 53.1±0.88s ? ?/sec 1.00 49.3±1.43s ? ?/sec
indexing/Indexing songs without faceted numbers 1.08 57.7±1.33s ? ?/sec 1.00 53.3±0.98s ? ?/sec
indexing/Indexing wiki 1.06 1051.1±21.46s ? ?/sec 1.00 989.6±24.55s ? ?/sec
indexing/Indexing wiki in three batches 1.20 1184.8±8.93s ? ?/sec 1.00 989.7±7.06s ? ?/sec
indexing/Reindexing geo_point 1.04 67.5±0.75s ? ?/sec 1.00 64.9±0.32s ? ?/sec
indexing/Reindexing movies with default settings 1.12 13.9±0.17s ? ?/sec 1.00 12.4±0.13s ? ?/sec
indexing/Reindexing songs with default settings 1.05 60.6±0.84s ? ?/sec 1.00 57.5±0.99s ? ?/sec
indexing/Reindexing wiki 1.07 1725.0±17.92s ? ?/sec 1.00 1611.4±9.90s ? ?/sec
```
Co-authored-by: Loïc Lecrenier <loic@meilisearch.com>
608: Fix soft deleted documents r=ManyTheFish a=ManyTheFish
When we replaced or updated some documents, the indexing was skipping the replaced documents.
Related to https://github.com/meilisearch/meilisearch/issues/2672
Co-authored-by: ManyTheFish <many@meilisearch.com>
594: Fix(Search): Fix phrase search candidates computation r=Kerollmops a=ManyTheFish
This bug is an old bug but was hidden by the proximity criterion,
Phrase searches were always returning an empty candidates list when the proximity criterion is deactivated.
Before the fix, we were trying to find any words[n] near words[n]
instead of finding any words[n] near words[n+1], for example:
for a phrase search '"Hello world"' we were searching for "hello" near "hello" first, instead of "hello" near "world".
Co-authored-by: ManyTheFish <many@meilisearch.com>
NOTE: The token_at_depth is method is a bit useless now, as the only
cases where there would be a toke at depth 1000 are the cases where
the parser already stack-overflowed earlier.
Example: (((((... (x=1) ...)))))
602: Use mimalloc as the default allocator r=Kerollmops a=loiclec
## What does this PR do?
Use mimalloc as the global allocator for milli's benchmarks on macOS.
## Why?
On Linux, we use jemalloc, which is a very fast allocator. But on macOS, we currently use the system allocator, which is very slow. In practice, this difference in allocator speed means that it is difficult to gain insight into milli's performance by running benchmarks locally on the Mac.
By using mimalloc, which is another excellent allocator, we reduce the speed difference between the two platforms.
Co-authored-by: Loïc Lecrenier <loic@meilisearch.com>
New full snapshot:
---
source: milli/src/update/word_prefix_pair_proximity_docids.rs
---
5 a 1 [101, ]
5 a 2 [101, ]
5 am 1 [101, ]
5 b 4 [101, ]
5 be 4 [101, ]
am a 3 [101, ]
amazing a 1 [100, ]
amazing a 2 [100, ]
amazing a 3 [100, ]
amazing an 1 [100, ]
amazing an 2 [100, ]
amazing b 2 [100, ]
amazing be 2 [100, ]
an a 1 [100, ]
an a 2 [100, 202, ]
an am 1 [100, ]
an an 2 [100, ]
an b 3 [100, ]
an be 3 [100, ]
and a 2 [100, ]
and a 3 [100, ]
and a 4 [100, ]
and am 2 [100, ]
and an 3 [100, ]
and b 1 [100, ]
and be 1 [100, ]
at a 1 [100, 202, ]
at a 2 [100, 101, ]
at a 3 [100, ]
at am 2 [100, 101, ]
at an 1 [100, 202, ]
at an 3 [100, ]
at b 3 [101, ]
at b 4 [100, ]
at be 3 [101, ]
at be 4 [100, ]
beautiful a 2 [100, ]
beautiful a 3 [100, ]
beautiful a 4 [100, ]
beautiful am 3 [100, ]
beautiful an 2 [100, ]
beautiful an 4 [100, ]
bell a 2 [101, ]
bell a 4 [101, ]
bell am 4 [101, ]
extraordinary a 2 [202, ]
extraordinary a 3 [202, ]
extraordinary an 2 [202, ]
house a 3 [100, 202, ]
house a 4 [100, 202, ]
house am 4 [100, ]
house an 3 [100, 202, ]
house b 2 [100, ]
house be 2 [100, ]
rings a 1 [101, ]
rings a 3 [101, ]
rings am 3 [101, ]
rings b 2 [101, ]
rings be 2 [101, ]
the a 3 [101, ]
the b 1 [101, ]
the be 1 [101, ]
