Skip to content
Permalink

Comparing changes

Choose two branches to see what’s changed or to start a new pull request. If you need to, you can also or learn more about diff comparisons.

Open a pull request

Create a new pull request by comparing changes across two branches. If you need to, you can also . Learn more about diff comparisons here.
base repository: ClickHouse/ClickBench
Failed to load repositories. Confirm that selected base ref is valid, then try again.
Loading
base: main
Choose a base ref
...
head repository: ClickHouse/ClickBench
Failed to load repositories. Confirm that selected head ref is valid, then try again.
Loading
compare: add-slothdb
Choose a head ref
Checking mergeability… Don’t worry, you can still create the pull request.
  • 4 commits
  • 24 files changed
  • 2 contributors

Commits on May 17, 2026

  1. Add SlothDB 

    SlothDB is an embedded analytical columnar SQL database
(https://github.com/SouravRoy-ETL/slothdb). Modeled the directory on
duckdb/: prebuilt x86-64 binary from the upstream GitHub release,
single-file .slothdb database, and one-shot `slothdb -c` invocations.
External wall-clock timing in ./query since SlothDB has no built-in
.timer.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
    @alexey-milovidov @claude
    alexey-milovidov and claude committed May 17, 2026
    Configuration menu
    Copy the full SHA
    3abfd48 View commit details
    Browse the repository at this point in the history

  2. Add SlothDB (Parquet, single) 

    Stateless variant: a VIEW over hits.parquet via read_parquet() instead
of ingesting the data. Modeled on duckdb-parquet/, with the same time
normalisation done inside the view so queries.sql matches the loaded
slothdb/ variant (no CREATE MACRO dependency).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
    @alexey-milovidov @claude
    alexey-milovidov and claude committed May 17, 2026
    Configuration menu
    Copy the full SHA
    addc25f View commit details
    Browse the repository at this point in the history

  3. slothdb: apt-install libgomp1 in ./install 

    The prebuilt linux-x64 binary links against libgomp at runtime, but
Ubuntu 24.04 cloud images don't ship it by default. Without the
package, `slothdb -c 'SELECT 1'` (i.e. ./check) immediately fails with
"error while loading shared libraries: libgomp.so.1" and bench_check_loop
burns the full 300s timeout before declaring the system unreachable.
Observed on the c6a.4xlarge run of the add-slothdb branch.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
    @alexey-milovidov @claude
    alexey-milovidov and claude committed May 17, 2026
    Configuration menu
    Copy the full SHA
    1676bd2 View commit details
    Browse the repository at this point in the history

  4. slothdb: drop unsupported SQL syntax from load + view 

    SlothDB's parser doesn't accept DuckDB's `SELECT * REPLACE (...)`
rewrite (REPLACE is only recognised inside CREATE OR REPLACE), and
it doesn't bind `read_parquet(..., binary_as_string=True)` keyword
args either — both forms produce `Parser Error: Unexpected token '('`
during load on the c6a.4xlarge run.

Switch to the form slothdb's own bench/run.py uses: CREATE TABLE /
CREATE VIEW AS SELECT * FROM 'hits.parquet'. SlothDB's parser
treats a bare string literal in FROM as an implicit table function
(`__FILE__`) that auto-detects the parquet format.

Side effect: time columns stay as integers rather than being
rewritten to DATE / TIMESTAMP, so Q19 and Q43 (extract(minute FROM
EventTime) and DATE_TRUNC('minute', EventTime)) may fail with a
type error. This matches slothdb upstream's own behaviour.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
    @alexey-milovidov @claude
    alexey-milovidov and claude committed May 17, 2026
    Configuration menu
    Copy the full SHA
    e4b76cf View commit details
    Browse the repository at this point in the history
Loading