array tree layouts - #8066
Conversation
Merging this PR will degrade performance by 19.91%
|
| Mode | Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|---|
| ❌ | Simulation | chunked_bool_canonical_into[(1000, 10)] |
20.4 µs | 35.3 µs | -42.15% |
| ❌ | Simulation | chunked_varbinview_into_canonical[(1000, 10)] |
177.2 µs | 213.3 µs | -16.94% |
| ❌ | Simulation | bitwise_not_vortex_buffer_mut[128] |
186.1 ns | 215.3 ns | -13.55% |
| ❌ | Simulation | chunked_varbinview_canonical_into[(100, 100)] |
273.8 µs | 308.6 µs | -11.3% |
| ❌ | Simulation | bitwise_not_vortex_buffer_mut[1024] |
246.4 ns | 275.6 ns | -10.58% |
Tip
Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.
Comparing os/array-tree-layout-stacked (7c9a72b) with os/array-serialise-columnar (eb11f18)
Footnotes
-
10 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩
93e517d to
3257531
Compare
39bfedf to
1a052bd
Compare
3257531 to
3c34669
Compare
1a052bd to
4636e34
Compare
c33a177 to
e40b94f
Compare
744833b to
9f963b1
Compare
e40b94f to
5739636
Compare
9f963b1 to
d4dc062
Compare
5739636 to
eb11f18
Compare
Signed-off-by: Onur Satici <onur@spiraldb.com>
Signed-off-by: Onur Satici <onur@spiraldb.com>
d4dc062 to
7c9a72b
Compare
|
This PR has been marked as stale because it has been open for 14 days with no activity. Please comment or remove the stale label if you wish to keep it active, otherwise it will be closed in 7 days |
|
This PR has been marked as stale because it has been open for 14 days with no activity. Please comment or remove the stale label if you wish to keep it active, otherwise it will be closed in 7 days |
Summary
Adds two new layouts, array tree layout and array tree flat layout. Latter writes the array node into a vortex array instead of serialising it into a flatbuffer, and the former collects all these columnar array nodes and creates one segment from them and writes at the end of the file, before the zone maps.
This replaces inlining the array nodes as flatbuffers to the footer