Influxdata
/
influxdb
mirror of https://github.com/influxdata/influxdb.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
influxdb/query_tests
History
Marco Neumann fecbbd9fa1
refactor: improve namespace caching in querier (#5492) 
1. Cache converted schema instead of catalog schema. This safes a buch
   of memcopies during conversion.
2. Simplify creation of new chunks, we now only need a `CachedTable`
   instead of a namespace and a table schema.

In an artificial benchmark, this removed around 10ms from the query
(although that was prior to #5467 which moved schema conversion one
level up). Still I think it is the cleaner cache design.

Co-authored-by: kodiakhq[bot] <49736102+kodiakhq[bot]@users.noreply.github.com>
 		2022-08-30 11:42:21 +00:00
..
cases chore: Update datafusion (with new sqlparser release) - option 1 (#5433) 2022-08-29 19:10:00 +00:00
generate fix: ensure querier cache is refreshed for partition sort key (#4660) 2022-05-25 10:44:42 +00:00
src refactor: improve namespace caching in querier (#5492) 2022-08-30 11:42:21 +00:00
Cargo.toml feat: refresh namespace cache before expiration (#5449) 2022-08-29 11:52:18 +00:00
README.md fix: Remove db crate from query_tests 2022-05-06 11:30:36 -04:00

README.md

This crate contains "integration" tests for the query engine. Specifically, it runs queries against a fully created database instance, records the output, and compares it to expected output.

Some tests simply have their inputs and outputs hard coded into #[test] annotated tests as is Rust's norm.

The tests in src/runner are driven somewhat more dynamically based on input files.

Cookbook: Adding a new Test

How to make a new test:

  1. Add a new file .sql to the cases/in directory
  2. Regenerate file: (cd generate && cargo run)
  3. Run the tests `` cargo test -p query_tests`
  4. You will get a failure message that contains examples of how to update the files

Example output

Possibly helpful commands:
  # See diff
  diff -du "/Users/alamb/Software/influxdb_iox/query_tests/cases/in/pushdown.expected" "/Users/alamb/Software/influxdb_iox/query_tests/cases/out/pushdown.out"
  # Update expected
  cp -f "/Users/alamb/Software/influxdb_iox/query_tests/cases/in/pushdown.out" "/Users/alamb/Software/influxdb_iox/query_tests/cases/out/pushdown.expected"

Cookbook: Adding a new test scenario

Each test can be defined in terms of a "setup" (a set of actions taken to prepare the state of database).

In the future, we envision more fine grained control of these setups (by implementing some of the database commands as IOX_TEST commands), but for now they are hard coded.

The SQL files refer to the setups with a specially formatted comment:

-- IOX_SETUP: OneMeasurementFourChunksWithDuplicates

To add a new setup, follow the pattern in scenario.rs of get_all_setups.