Update Ktor to 3.6.0

[bjhham]

Description

This updates Ktor to the latest version, including HTTP/3 and a number of performance improvements that I introduced after hammering away at the Netty engine a couple weeks ago.

---

PR Commands — comment on this PR to trigger (requires collaborator approval):

Command	Description
/benchmark -f <framework>	Run every test the framework subscribes to
/benchmark -f <framework> -t <test>	Run one test only
/benchmark -f <framework> --save	Run and save results (updates the leaderboard on merge)
/benchmark -f <framework> -t <test> --save	Run one test and save results
/benchmark -f <framework> --compare <other>	Measure the deltas against another framework instead of this one
/benchmark-multiple -f <fw1>,<fw2>,...	Benchmark several frameworks in one run — takes -t and --save too; saved results land in a single commit
/benchmark-multiple --save	No -f needed: benchmark and save every framework the PR touches
/benchmark-test -t <test>	Benchmark all enabled frameworks subscribed to <test> and save the results

For /benchmark, always specify -f <framework>; the flags combine in any order. Results come back as a comment with a per-profile table of RPS, p99, CPU and memory — one table per framework on multi runs. A new benchmark comment while a run is in flight queues behind it (one deep) instead of cancelling it. For multi-framework PRs (dependency bumps, same-language refactors) prefer /benchmark-multiple, which runs everything in a single job and commits all saved results together, so no run overwrites another. --compare works on single-framework runs only.

What the deltas are measured against. By default, this framework's own results published on main - answering "did this change help?". When you are tuning a variant or a successor entry, --compare re-bases them on another entry instead:

/benchmark -f genhttp --compare genhttp-kestrel

The reply states which baseline it used, and profiles the other framework does not run show n/a rather than a delta.

---

Run benchmarks locally

You can validate and benchmark your framework locally with the lite script — no CPU pinning, fixed connection counts, all load generators run in Docker.

./scripts/validate.sh <framework>
./scripts/benchmark-lite.sh <framework> baseline
./scripts/benchmark-lite.sh --load-threads 4 <framework>

Requirements: Docker Engine on Linux. Load generators (gcannon, h2load, h2load-h3, wrk) are built as self-contained Docker images on first run.

[bjhham]

/benchmark-multiple -f ktor,ktor-tuned

[github-actions]

👋 Benchmark request received. A collaborator will review and approve the run.

[github-actions]

Benchmark Results

Frameworks: 2 | Test: all tests

✅ ktor

Test	Conn	RPS	Rate	CPU/req	p99	p99.9	CPU	Mem	Δ RPS	Δ Mem
baseline	4096	956,018	—	—	—	—	5777.2%	5.2GiB	+16.3%	-45.3%
pipelined	4096	2,112,308	—	—	—	—	5401.4%	9.2GiB	+261.2%	-48.6%
limited-conn	4096	709,415	—	—	—	—	5537.8%	5.7GiB	+6.6%	+29.5%
json-comp	4096	203,187	—	—	—	—	6062.1%	5.5GiB	+5.3%	+5.8%
json-comp	16384	203,258	—	—	—	—	6239.1%	9.9GiB	+7.3%	-11.6%
json-tls	4096	482,107	—	—	—	—	4763.1%	8.1GiB	+23.0%	-12.9%
8gbit	512	49,336	0.9867	116.78us	245.0us	12700.0us	543.1%	1.5GiB	+0.3%	-44.4%
static-tls	1024	199,936	—	—	—	—	6325.0%	4.4GiB	-2.3%	+15.8%
async-db	1024	48,494	—	—	—	—	2504.1%	2.9GiB	-53.3%	-69.8%
fortunes	1024	62,105	—	—	—	—	5033.5%	10.7GiB	+21.7%	+42.7%
baseline-h2	256	804,286	—	—	—	—	6299.7%	14.4GiB	+44.0%	-51.7%
baseline-h2	1024	697,599	—	—	—	—	6351.7%	24.7GiB	+28.7%	-17.7%
static-h2	256	211,874	—	—	—	—	5969.1%	9.0GiB	+11.7%	-43.8%
static-h2	1024	134,286	—	—	—	—	5720.3%	9.5GiB	-25.5%	-61.5%
baseline-h2c	256	838,087	—	—	—	—	6438.5%	12.7GiB	+103.3%	-58.6%
baseline-h2c	1024	759,599	—	—	—	—	6357.6%	24.8GiB	+88.5%	-20.3%
baseline-h2c	4096	452,286	—	—	—	—	5680.4%	31.2GiB	+61.5%	~0%
json-h2c	1024	447,200	—	—	—	—	5438.9%	12.6GiB	+42.3%	-55.3%
json-h2c	4096	419,016	—	—	—	—	5342.3%	27.4GiB	+48.2%	-8.4%
baseline-h3	64	338,472	—	—	—	—	4629.4%	3.2GiB	NEW	NEW
static-h3	64	102,955	—	—	—	—	5888.6%	2.4GiB	NEW	NEW
echo-ws	512	1,024,145	—	—	—	—	3411.5%	2.8GiB	+0.3%	+12.0%
echo-ws	4096	950,673	—	—	—	—	3379.0%	2.0GiB	-3.0%	-69.7%
echo-ws	16384	783,696	—	—	—	—	3287.8%	6.1GiB	+21.4%	-58.8%
echo-ws-pipeline	512	11,172,181	—	—	—	—	3360.2%	3.0GiB	-0.4%	-72.2%
echo-ws-pipeline	4096	10,628,080	—	—	—	—	3400.9%	3.0GiB	-0.2%	-76.9%
echo-ws-pipeline	16384	8,054,771	—	—	—	—	3475.1%	6.3GiB	-9.3%	-69.6%
latency-1m	1024	997,582	0.9976	53.22us	16872.0us	128288.0us	5396.6%	4.9GiB	+13.5%	-2.0%
latency-10k	1024	9,981	0.9981	75.12us	152.0us	2219.0us	72.1%	1.5GiB	~0%	+230.3%
latency-500k-8cpu	1024	210,364	0.4207 ⚠️	32.85us	11423744.0us	11702272.0us	682.1%	1.3GiB	+6.6%	~0%
async	32000	226,870	—	—	—	—	2338.1%	6.7GiB	-7.1%	-59.9%

✅ ktor-tuned

Test	Conn	RPS	Rate	CPU/req	p99	p99.9	CPU	Mem	Δ RPS	Δ Mem
baseline	4096	1,281,059	—	—	—	—	6518.6%	9.8GiB	~0%	~0%
pipelined	4096	2,437,398	—	—	—	—	6070.3%	9.8GiB	~0%	~0%
limited-conn	4096	547,139	—	—	—	—	4674.8%	9.8GiB	~0%	~0%
json-comp	4096	240,495	—	—	—	—	6245.3%	6.6GiB	~0%	~0%
json-comp	16384	236,191	—	—	—	—	6092.6%	9.4GiB	~0%	~0%
json-tls	4096	842,743	—	—	—	—	6324.2%	6.8GiB	~0%	~0%
8gbit	512	49,329	0.9866	87.54us	191.0us	10140.0us	471.6%	1.7GiB	~0%	~0%
static-tls	1024	258,353	—	—	—	—	6310.6%	4.6GiB	~0%	~0%
async-db	1024	52,833	—	—	—	—	2227.2%	2.8GiB	~0%	~0%
baseline-h2	256	1,118,369	—	—	—	—	6337.5%	10.6GiB	~0%	~0%
baseline-h2	1024	1,048,754	—	—	—	—	6360.2%	26.0GiB	~0%	~0%
static-h2	256	227,851	—	—	—	—	6196.7%	8.6GiB	~0%	~0%
static-h2	1024	213,239	—	—	—	—	5841.5%	12.0GiB	~0%	~0%
baseline-h2c	256	1,162,718	—	—	—	—	6344.3%	13.7GiB	~0%	~0%
baseline-h2c	1024	1,086,015	—	—	—	—	6427.4%	25.9GiB	~0%	~0%
baseline-h2c	4096	535,627	—	—	—	—	5128.5%	29.9GiB	~0%	~0%
json-h2c	1024	705,685	—	—	—	—	6238.9%	12.0GiB	~0%	~0%
json-h2c	4096	665,557	—	—	—	—	6143.6%	26.3GiB	~0%	~0%
baseline-h3	64	362,107	—	—	—	—	2926.8%	4.8GiB	~0%	~0%
static-h3	64	91,646	—	—	—	—	3500.8%	2.1GiB	~0%	~0%
echo-ws	512	994,108	—	—	—	—	6292.8%	1.7GiB	~0%	~0%
echo-ws	4096	1,135,731	—	—	—	—	6359.0%	4.3GiB	~0%	~0%
echo-ws	16384	990,277	—	—	—	—	6263.9%	6.3GiB	~0%	~0%
echo-ws-pipeline	512	11,381,851	—	—	—	—	6194.0%	8.7GiB	~0%	~0%
echo-ws-pipeline	4096	12,543,004	—	—	—	—	6282.9%	11.0GiB	~0%	~0%
echo-ws-pipeline	16384	12,471,290	—	—	—	—	6270.1%	10.9GiB	~0%	~0%
latency-1m	1024	997,578	0.9976	54.94us	3145.0us	12724.0us	5893.5%	5.2GiB	~0%	~0%
latency-10k	1024	9,981	0.9981	45.64us	131.0us	2107.0us	43.3%	1.6GiB	~0%	~0%
latency-500k-8cpu	1024	265,880	0.5318 ⚠️	30.00us	9383936.0us	9728000.0us	790.9%	2.4GiB	~0%	~0%
async	32000	316,403	—	—	—	—	2381.6%	7.5GiB	~0%	~0%