From 1b7a6687c8b7b6d3584bdeea94f6d65e69c8641c Mon Sep 17 00:00:00 2001 From: karan bamal <51521112+abankalarm@users.noreply.github.com> Date: Fri, 24 Apr 2020 13:41:27 +0530 Subject: [PATCH] Update README.md (#630) MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit * Update README.md * Update README.md Co-Authored-By: Clément Renault Co-authored-by: Clément Renault --- README.md | 34 +--------------------------------- 1 file changed, 1 insertion(+), 33 deletions(-) diff --git a/README.md b/README.md index 526777cc5..d97df08f8 100644 --- a/README.md +++ b/README.md @@ -163,40 +163,8 @@ You can access the web interface in your web browser at the root of the server. ### Documentation -Now that your MeiliSearch server is up and running, you can learn more about how to tune your search engine in [the documentation](https://docs.meilisearch.com). +Now that your MeiliSearch server is up and running, you can learn more about how to tune your search engine in [the documentation](https://docs.meilisearch.com). -### Technical features - -- Provides [6 default ranking criteria](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-core/src/criterion/mod.rs#L106-L111) used to [bucket sort](https://en.wikipedia.org/wiki/Bucket_sort) documents -- Accepts [custom criteria](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-core/src/criterion/mod.rs#L20-L29) and can apply them in any custom order -- Supports [ranged queries](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-core/src/query_builder.rs#L342), useful for paginating results -- Can [distinct](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-core/src/query_builder.rs#L324-L329) and [filter](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-core/src/query_builder.rs#L313-L318) returned documents based on context defined rules -- Searches for [concatenated](https://github.com/meilisearch/MeiliSearch/pull/164) and [splitted query words](https://github.com/meilisearch/MeiliSearch/pull/232) to improve the search quality. -- Can store complete documents or only [user schema specified fields](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/datasets/movies/schema.toml) -- The [default tokenizer](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-tokenizer/src/lib.rs) can index Latin based languages and Kanji characters -- Returns [the matching text areas](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-types/src/lib.rs#L49-L65), useful to highlight matched words in results -- Accepts query time search config like the [searchable attributes](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-core/src/query_builder.rs#L331-L336) -- Supports [runtime incremental indexing](https://github.com/meilisearch/MeiliSearch/blob/3ea5aa18a209b6973b921542d46a79e1c753c163/meilisearch-core/src/store/mod.rs#L143-L212) - -## Performance - -When processing a dataset composed of 5M books, each with their own titles and authors, MeiliSearch is able to carry out more than 553 req/sec with an average response time of 21 ms on an Intel i7-7700 (8) @ 4.2GHz. - -Requests are made using [wrk](https://github.com/wg/wrk) and scripted to simulate real users' queries. - -``` -Running 10s test @ http://1.2.3.4:7700 - 2 threads and 10 connections - Thread Stats Avg Stdev Max +/- Stdev - Latency 21.45ms 15.64ms 214.10ms 85.95% - Req/Sec 256.48 37.66 330.00 69.50% - 5132 requests in 10.05s, 2.31MB read -Requests/sec: 510.46 -Transfer/sec: 234.77KB -``` - -We also indexed a dataset containing about _12 millions_ cities names in _24 minutes_ on a _8 cores_, _64 GB of RAM_, and a _300 GB NMVe_ SSD machine.
-The size of the resulting database reached _16 GB_ and search results were presented between _30 ms_ and _4 seconds_ for short prefix queries. ## Contributing