Add System.Diagnostics.TextMapPropagator support

[shirhatti]

No description provided.

[shirhatti]

Plaintext on aspnet-citrine-lin

application	old	new
CPU Usage (%)	99	99	0.00%
Cores usage (%)	2,781	2,778	-0.11%
Working Set (MB)	415	415	0.00%
Private Memory (MB)	1,006	1,557	+54.77%
Build Time (ms)	3,553	5,432	+52.88%
Start Time (ms)	150	171	+14.00%
Published Size (KB)	106,173	106,173	0.00%
.NET Core SDK Version	6.0.100-preview.6.21276.12	6.0.100-preview.6.21276.12
Max CPU Usage (%)	181	99	-45.30%
Max Working Set (MB)	434	435	+0.23%
Max GC Heap Size (MB)	317	302	-4.73%
Size of committed memory by the GC (MB)	363	362	-0.28%
Max Number of Gen 0 GCs / sec	7.00	7.00	0.00%
Max Number of Gen 1 GCs / sec	1.00	1.00	0.00%
Max Number of Gen 2 GCs / sec	1.00	1.00	0.00%
Max Time in GC (%)	1.00	0.00
Max Gen 0 Size (B)	672	672	0.00%
Max Gen 1 Size (B)	4,125,536	1,060,016	-74.31%
Max Gen 2 Size (B)	3,181,976	3,221,496	+1.24%
Max LOH Size (B)	230,160	672	-99.71%
Max Allocation Rate (B/sec)	2,040,420,720	2,040,749,408	+0.02%
Max GC Heap Fragmentation	1	1	-9.88%
# of Assemblies Loaded	94	93	-1.06%
Max Exceptions (#/s)	697	687	-1.43%
Max Lock Contention (#/s)	263	361	+37.26%
Max ThreadPool Threads Count	48	48	0.00%
Max ThreadPool Queue Length	207	185	-10.63%
Max ThreadPool Items (#/s)	176,635	186,574	+5.63%
Max Active Timers	0	0
IL Jitted (B)	172,641	183,274	+6.16%
Methods Jitted	1,859	2,072	+11.46%

load	old	new
CPU Usage (%)	19	19	0.00%
Cores usage (%)	531	531	0.00%
Working Set (MB)	37	37	0.00%
Private Memory (MB)	358	358	0.00%
Start Time (ms)	0	0
First Request (ms)	70	72	+2.86%
Requests/sec	2,476,644	2,475,423	-0.05%
Requests	37,394,952	37,378,301	-0.04%
Mean latency (ms)	1.15	1.15	0.00%
Max latency (ms)	49.90	53.67	+7.56%
Bad responses	0	0
Socket errors	0	0
Read throughput (MB/s)	311.77	311.62	-0.05%
Latency 50th (ms)	0.99	0.96	-3.03%
Latency 75th (ms)	1.46	1.41	-3.42%
Latency 90th (ms)	2.07	1.91	-7.73%
Latency 99th (ms)	0.00	0.00

[shirhatti]

@Tratcher The only new change I've made since we last spoke is to resolve the Propagator from DI instead of using the global static. Can you review that?

Also, the difference in the GC counters makes no sense to me 🤷. Can you help me make sense of that?

[Tratcher]

The DI change seems fine

[Tratcher]

No idea on the GC data, but at least it's getting better and not worse 😁.

[shirhatti]

There's no way LOH is 672 bytes. Kestrel's slabbed memory pool is allocated in the LOH. @sebastienros any ideas? Repeating the run, I got 131800 for Max LOH size for both before/after.

As for reduction in Max Gen 1 and increase in Max Gen 2 size, I'm not sure what to make of it? Something is getting promoted sooner? I'll try collecting an allocation profile to make more sense of this

[tarekgh]

    private static HostingApplication CreateApplication(IHttpContextFactory httpContextFactory = null, bool useHttpContextAccessor = false,

May be a clever idea in a future PR to add a test consuming other propagator. e.g., tests create a custom propagator and consume it. You don't have to block this PR on that.

---

Refers to: src/Hosting/Hosting/test/HostingApplicationTests.cs:182 in abcc758. [](commit_id = abcc758, deletion_comment = False)

[tarekgh]

Added some commented. LGTM otherwise.

[Tratcher]

You used to add these in reverse, does ExtractBaggage handle the reversal? If so then include that in the comment. Otherwise this comment about order sensitivity doesn't make sense.

[shirhatti]

I’ll send a follow-up PR to fix the comment. I want to merge this when the PR checks are green in case AzDo has a bad day again

[davidfowl]

I'm assuming there's no extra allocations here.