shoney.arickathil
6183a67dfc
databasev2 4 part A, task 5. Controlled before/after — same machine, same workload (wmix 4000 32), same build except db.c and vm.c, two runs each interleaved: - per-statement barrier: 2213 / 2177 ops/sec, p50 7183 / 7251us - group commit: 6216 / 6525 ops/sec, p50 3458 / 3444us - ~2.9x throughput, ~2.1x lower p50 The full campaign confirms it a second way: s1 takes the inline path and commits per statement BY DESIGN, so within one build the shard configs are batching-off vs batching-on — 1467 -> 5117 ops/sec, mean batch 1.0 -> 5.43, peak 1 -> 57. 3.5x, agreeing with the 2.9x above. Recorded honestly: - the BEFORE p99 is at the histogram ceiling (hist_add clamps at 20000us and both runs pinned there), so the true figure is >=20ms and unknown. The improvement is AT LEAST 2.3x; the old p99 was off the instrument - durable.sN.mixwrite went 480 -> 492 ops/sec, i.e. UNCHANGED. That was the spec's original payoff metric and correcting it was part of the brainstorm: mix performs 20 writes at C=4, mean batch 1.01. A workload that never has two writes in flight cannot be helped by batching them - seed is likewise unchanged: a serial writer has nothing to batch with - so the payoff is real but CONDITIONAL — it appears where concurrent durable writes fan into the owner shard, and nowhere else Two traps recorded in perf-targets §6: - do not benchmark durability on /tmp: it is tmpfs here, where fdatasync is free. The same run reported 195000 ops/sec at p50 1us there against 2200 at p50 7200us on ext4 — no barrier to amortise, so the measurement measures nothing. db-bench keeps its stores under bench/ for this reason - the record count is not the update count: 7755 records for 4000 updates, because hist_dump and the done-marker are themselves durable inserts - FIXED a regression I introduced in T4: master's committed baseline is FULL mode (N=20000, crash_reps=3, msg_n=200000) and I had overwritten it with quick-mode values. Regenerated from a full campaign; the full run now passes 106 checks 0 failures against it - gate still bites: sN wmix ops_sec -70% -> FAIL on exactly that metric Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
140 lines
7 KiB
Markdown
140 lines
7 KiB
Markdown
# Performance targets — measured, named, waiting
|
||
|
||
A register like [`discarded.md`](discarded.md)/[`learnings.md`](learnings.md):
|
||
optimization candidates that exist because a NUMBER says so, not a
|
||
hunch. Every row cites its measurement (the db-bench campaign,
|
||
`bench/baseline.json`, or the [go-sqlite comparison](../../bench/compare/go-sqlite/README.md))
|
||
and names an owner iteration when one exists. A target leaves this file
|
||
by landing (delta recorded in the baseline) or by being rejected into
|
||
`discarded.md` with its reason.
|
||
|
||
## 1. The write path — update-through-query re-probes, insert re-encodes
|
||
|
||
**Measured 2026-08-22** (go-sqlite comparison, N=20k, same machine,
|
||
ext4): ram mixed writes 195,465 ops/s vs SQLite's 380,069 (×1.9
|
||
behind); durable mixed writes 2,324 vs 3,257 (×1.4 behind) — while
|
||
writeonce WINS durable seed ×1.4 and reads ×2.6–6.4. The write gap is
|
||
specifically the UPDATE half of the mix.
|
||
|
||
Suspected costs, in probable order (attribute before optimizing — the
|
||
C-API microbench the 22 spec reserves exists for exactly this):
|
||
|
||
1. **Update-through-query runs a whole query statement per update**:
|
||
probe (now O(1)) + materialize an id `multi` (arena alloc) +
|
||
`DB_GET_FIELD`/`DB_UPDATE_FIELD` builtin round-trips per touched
|
||
field. SQLite's equivalent is one page write inside one statement.
|
||
2. **`wo_row_update_field` walks every index three times** (shadow
|
||
unique check, old-entry removal, new-entry add — three
|
||
`touches`-loops over `t->indexes` per update; see
|
||
`database/src/table.c`).
|
||
3. **Insert encodes per field with a malloc per text/owned value**
|
||
(`db_val_encode`) — visible as ram seed ×1.2 behind SQLite (245k vs
|
||
297k) even though the durable flavor wins.
|
||
4. **A WAL update record re-encodes the whole row**
|
||
(`wo_wal_append_update` writes the row image, not a delta).
|
||
|
||
**Owner:** none yet. Sequence note: iteration 23 (io_uring
|
||
group-commit) rewrites the durable write path's syscall story anyway —
|
||
re-measure after 23 lands, then decide whether the RAM-side costs
|
||
(1–3) earn their own slice. Acceptance shape: ram write ops/s closes
|
||
on SQLite's number with reads unharmed; baseline refreshed with the
|
||
delta recorded.
|
||
|
||
## 2. Cross-shard DB RPC halves concurrent read throughput
|
||
|
||
**Measured 2026-08-22** (db-bench campaign): ram mixread 89,538 ops/s
|
||
single-shard vs 44,918 at default cores; durable 9,211 vs 4,324. The
|
||
DB actor serializes every statement on shard 0 and each op pays an
|
||
envelope + park/unpark round-trip.
|
||
|
||
**Owner: by design, priced deliberately** (story 8's settled decision
|
||
1 — rejected alternatives: engine lock, partitioned tables "wait for a
|
||
measured need"). THIS is the measured need's first data point; the
|
||
recorded escalation path is partitioned/replicated read state, only if
|
||
a real workload (iteration 24's chat) hurts. Not actionable before 24.
|
||
|
||
## 3. The mutex inbox costs ~6× on cross-shard message rate
|
||
|
||
**Measured 2026-08-22**: 16.7M msgs/s same-heap vs 2.85M cross-shard
|
||
(`msgrate`). **Owner: iteration 31** (mailbox/backpressure decisions
|
||
consume this number) and stage-2 deviation 4 (lock-free rings arrive
|
||
only if the mutex is the measured bottleneck — at 2.85M msgs/s it is
|
||
not the limiting factor for any current workload).
|
||
|
||
## 4. Durable write throughput is fsync-bound at ~4.5k/s
|
||
|
||
**Measured 2026-08-21**: durable seed 4,460 inserts/s vs ram 245k —
|
||
the ~55× gap is one fdatasync per statement (~220µs each).
|
||
**Owner: iteration 23** (io_uring group-commit) — its acceptance is
|
||
literally this number moving while the crash battery stays green.
|
||
|
||
## 6. WAL group commit: one barrier per drain (databasev2 4 part A)
|
||
|
||
**Measured 2026-08-28.** Before this, the engine committed per *statement*:
|
||
`db.c` called `wo_wal_commit` immediately after every append, so each row
|
||
change bought its own `pwrite` + `fdatasync`. Now shard 0 stages every queued
|
||
write request, issues one barrier, and only then releases the held replies.
|
||
|
||
### The controlled before/after
|
||
|
||
Same machine, same workload (`wmix 4000 32` — every op a durable update, 32
|
||
concurrent), same build except `db.c` and `vm.c`, two runs each, interleaved:
|
||
|
||
| | ops/sec | p50 | p99 |
|
||
| --- | --- | --- | --- |
|
||
| per-statement barrier | 2213 · 2177 | 7183 · 7251 µs | **20000 · 20000 µs** |
|
||
| group commit | **6216 · 6525** | **3458 · 3444 µs** | 11139 · 5971 µs |
|
||
|
||
**≈2.9× throughput, ≈2.1× lower p50.**
|
||
|
||
**The p99 "before" figure is at the histogram ceiling, not a measurement.**
|
||
`hist_add` clamps at 20000 µs, and both before-runs pinned there — so the true
|
||
before p99 is ≥20 ms and unknown. The improvement is *at least* 2.3×; the
|
||
honest statement is that the old p99 was off the end of the instrument.
|
||
|
||
### Confirmation from the committed baseline
|
||
|
||
The full campaign gives the same answer a second way. `s1` takes the inline
|
||
path, which commits per statement **by design**, so within one build the two
|
||
shard configurations are batching-off against batching-on:
|
||
|
||
| Leg | ops/sec | p50 | p99 | mean batch | peak batch |
|
||
| --- | --- | --- | --- | --- | --- |
|
||
| `durable.s1.wmix` (inline, unbatched) | 1467 | 455 µs | 721 µs | **1.0** | 1 |
|
||
| `durable.sN.wmix` (batched) | **5117** | 8208 µs | 12169 µs | **5.43** | 57 |
|
||
|
||
3.5× throughput, agreeing with the 2.9× above. Note `sN` latency is *higher*
|
||
while throughput is 3.5× better: 64 writers queueing behind one owner shard
|
||
trade per-op latency for barrier amortisation, which is what group commit is.
|
||
|
||
Batching scales with write concurrency exactly as designed — mean batch at
|
||
C = 4 / 16 / 64 was **1.13 / 1.76 / 5.35**, peak **3 / 10 / 39**.
|
||
|
||
### What did NOT improve, and why that was predicted
|
||
|
||
`durable.sN.mixwrite` went **480 → 492 ops/s** — unchanged. That is the metric
|
||
the spec *originally* named as the payoff, and correcting it was part of the
|
||
brainstorm: `mix` writes on one op in ten with C=4, so a quick run performs
|
||
**20 writes** and mean batch measured **1.01** over 3112 barriers. A workload
|
||
that never has two writes in flight cannot be helped by batching them.
|
||
`durable.*.seed` is likewise unchanged: a serial single writer has nothing to
|
||
batch with under any scheme.
|
||
|
||
**So the payoff is real but conditional: it appears exactly where concurrent
|
||
durable writes fan into the owner shard, and nowhere else.**
|
||
|
||
### Two traps worth recording
|
||
|
||
**Do not benchmark durability on `/tmp`.** It is `tmpfs` here, where
|
||
`fdatasync` is free — the same `wmix` run reported **195 000 ops/s at p50 1 µs**
|
||
there against **2200 ops/s at p50 7200 µs** on ext4. There is no barrier to
|
||
amortise on a memory filesystem, so a group-commit measurement taken there
|
||
measures nothing. `db-bench` gets this right by keeping its stores under
|
||
`bench/`.
|
||
|
||
**The record count is not the update count.** `wmix` staged 7755 records for
|
||
4000 updates because the histogram dump and the done-marker are themselves
|
||
durable inserts. They arrive as an end-of-run burst, which is batch-friendly,
|
||
so `mean_batch` is not purely update-driven. Peak staged bytes stayed small
|
||
(2793 B at C=64), which is what settled the decision to ship **no batch cap**:
|
||
the request queue's existing upstream bound is sufficient.
|