diff options
| author | eunhhu <[email protected]> | 2026-09-30 08:54:29 +0100 |
|---|---|---|
| committer | eunhhu <[email protected]> | 2026-09-30 08:54:29 +0100 |
| commit | 9840d27a9e437de32256024a2fd6a3219a239815 (patch) | |
| tree | eba57012bcdaac8e019d1421370a570a3e9f70d9 /research/experiments/022-tpcc-grouped | |
| parent | 987d4d7aa08624e28d0ea18a1e66e359a0230c80 (diff) | |
| download | pure-9840d27a9e437de32256024a2fd6a3219a239815.tar.gz pure-9840d27a9e437de32256024a2fd6a3219a239815.zip | |
Say what snapshots cost the owner, and what reusing their nodes gave (107)
Most of it was the allocator: leaves past glibc's thread cache, freed by
whichever thread let the last snapshot go. jemalloc was 11-14% faster
at more memory; reusing the nodes in Pure gave a median +9.4% over six
pairs, 4-8% less CPU a transaction in all six, and 1.97x the state bytes
resident where it was 2.18x. TPC-C in RAM is now 4.0x a tuned SQLite at
W=1 and 3.5x at W=4; finding items is the owner's largest part now.
Diffstat (limited to 'research/experiments/022-tpcc-grouped')
| -rw-r--r-- | research/experiments/022-tpcc-grouped/tables.py | 5 |
1 files changed, 3 insertions, 2 deletions
diff --git a/research/experiments/022-tpcc-grouped/tables.py b/research/experiments/022-tpcc-grouped/tables.py index 2c018b0..fcfd468 100644 --- a/research/experiments/022-tpcc-grouped/tables.py +++ b/research/experiments/022-tpcc-grouped/tables.py @@ -69,5 +69,6 @@ def table(name, device): print() -for name in sys.argv[1:] or ['ram-w1', 'ram-w4', 'sd-w1-*', 'sd-w4-*']: - table(name, name.startswith('sd')) +if __name__ == '__main__': + for name in sys.argv[1:] or ['ram-w1', 'ram-w4', 'sd-w1-*', 'sd-w4-*']: + table(name, name.startswith('sd')) |