Skip to content

Latest commit

 

History

History
103 lines (75 loc) · 4.69 KB

description.md

File metadata and controls

103 lines (75 loc) · 4.69 KB

English-language Wikipedia:Database page abstracts

Example Document

This use case generates 5.9 million docs, with 3 Text fields per document. On average each added document will have a size of 227 bytes.

(prior field tokenization)

{
'title': 'Wikipedia: Politics of the Democratic Republic of the Congo',
'url': 'https://en.wikipedia.org/wiki/Politics_of_the_Democratic_Republic_of_the_Congo',
'abstract': 'Politics of the Democratic Republic of Congo take place in a framework of a republic in transition from a civil war to a semi-presidential republic.'
}

Query types

Full text search queries

Query type Description Example Status
simple-1word-query Simple 1 Word Query Abraham ✔️
2word-union-query 2 Word Union Query Abraham Lincoln ✔️
2word-intersection-query 2 Word Intersection Query Abraham|Lincoln ✔️
exact-3word-match Exact 3 Word Match "President Abraham Lincoln" ✖️
autocomplete-1100-top3 Autocomplete -1100 Top 2-3 Letter Prefixes ✖️
2field-2word-intersection-query 2 Fields, one word each, Intersection query @text_field1: text_value1 @text_field2: text_value2 ✖️
2field-1word-intersection-1numeric-range-query 2 Fields, one text and another numeric, Intersection and numeric range query @text_field: text_value @numeric_field:[{min} {max}] ✖️

Spell Check queries

Performs spelling correction on a query, returning suggestions for misspelled terms. To simmulate misspelled terms, for each word a deterministic random number of edits in the range 0..Min(word.length/2 , 4) is chosen.

For each edit a random type of edit (delete, insert random char, replace with random char, switch adjacent chars).

Query type Description Example Status
simple-1word-spellcheck Simple 1 Word Spell Check Query FT.SPELLCHECK {index} reids DISTANCE 1 ✔️

Autocomplete queries

Query type Description Example Status
`` ✖️

Aggregate queries

Aggregations are a way to process the results of a search query, group, sort and transform them - and extract analytic insights from them. Much like aggregation queries in other databases and search engines, they can be used to create analytics reports, or perform Faceted Search style queries.

Query type Description Clauses included Status
`` ✖️

Synonym queries

Query type Description Example Status
`` ✖️

How to benchmark

Using FTSB for benchmarking involves 2 phases: data and query generation, and query execution.
The following steps focus on how to retrieve the data and generate the commands for the nyc_taxis use case.

Generating the dataset

To generate the required dataset command file issue:

cd $GOPATH/src/github.com/RediSearch/ftsb/scripts/datagen_redisearch/enwiki_abstract
python3 ftsb_generate_enwiki_abstract.py

Index properties

The use case generates an secondary index with 3 fields per document:

  • 3 TEXT sortable fields.

Running the benchmark

Assuming you have redisbench-admin and ftsb_redisearch installed, for the default dataset, run:

redisbench-admin run \
     --repetitions 3 \
     --benchmark-config-file  https://s3.amazonaws.com/benchmarks.redislabs/redisearch/datasets/enwiki_abstract-hashes/enwiki_abstract-hashes.redisearch.cfg.json

Key Metrics:

After running the benchmark you should have a result json file generated, containing key information about the benchmark run(s). Focusing specifically on this benchmark the following metrics should be taken into account and will be used to automatically choose the best run and assess results variance, ordered by the following priority ( in case of results comparison ):

Setup step key metrics

Metric Family Metric Name Unit Comparison mode
Throughput Overall Ingestion rate docs/sec higher is better
Latency Overall ingestion p50 milliseconds lower is better

Benchmark step key metrics

Metric Family Metric Name Unit Comparison mode
Throughput Overall Updates and Aggregates query rate docs/sec Higher is better
Latency Overall Updates and Aggregates query q50 latency milliseconds Lower is better