Change to the NetBSD archive URL rather than the CDN

[tgross35]

The CDN has been down for a few hours. Switch to an alternative for the time being so we can unblock CI.

It appears that the CDN is quite a bit faster, so we will likely want to switch back when available.

Context: https://rust-lang.zulipchat.com/#narrow/stream/242791-t-infra/topic/dist-x86_64-netbsd.20is.20borked

try-job: dist-x86_64-netbsd

[rustbot]

r? @Kobzol

rustbot has assigned @Kobzol.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[Noratrieb]

we should not do that, we should instead mirror it ourselves and use that URL.

[tgross35]

Right, I updated the fixme to say that

[tgross35]

Just realized the archive URL uses http and doesn't seem to have https, we may not want this then :/

I'll give it another half hour to see if the CDN comes back. If not, I'll add a SHA sum check. Nothing to calculate the correct SHA from except what's currently available at the archive URL, unfortunately... Open to suggestions

[notriddle]

Nothing to calculate the correct SHA from except what's currently available at the archive URL, unfortunately... Open to suggestions

https://cdn.netbsd.org/pub/NetBSD/security/hashes/NetBSD-9.0_hashes.asc seems to work

[tgross35]

CDN is still down. I added a check against the sha from @notriddle's link.

Could somebody @bors try if this looks correct?

[Noratrieb]

@bors try

[bors]

⌛ Trying commit 95b682c with merge cf0c6be...

[tgross35]

Why no dist-x86_64-netbsd in the try, it's in the top post 🤔

[tgross35]

Does <!-- homu-ignore:start --> .. <!-- homu-ignore:end --> make bors ignore try-job directives too? (my try-job was wrapped in those). Thought that was just what gets stripped from the commit message. Removed those comments in case that's it (I guess @Nilstrieb could you retry the try?)

edit: looks like this shouldn't matter?

rust/src/ci/github-actions/calculate-job-matrix.py

Lines 74 to 87 in 7d97c59

	def get_custom_jobs(ctx: GitHubCtx) -> List[str]:
	"""
	Tries to parse names of specific CI jobs that should be executed in the form of
	try-job: <job-name>
	from the commit message of the passed GitHub context.
	"""
	if ctx.commit_message is None:
	return []
	regex = re.compile(r"^try-job: (.*)", re.MULTILINE)
	jobs = []
	for match in regex.finditer(ctx.commit_message):
	jobs.append(match.group(1))
	return jobs

. No clue what happened then

edit edit: unless it needs to be in the actual commit message?

[Kobzol]

I think that the problem is that the stuff inside the comments is indeed ignored by homu, and it then doesn't add it to the commit message. You can see that it is not here: 95b682c.

CI reads the commit message, which is generated from the PR description by homu, but here it was ignored.

[tgross35]

Oh that makes more sense, thanks. I removed the comments, mind restarting the try? (assuming this still seems necessary)

[DianQK]

@bors try

[bors]

⌛ Trying commit 95b682c with merge 5ec9064...

[bors]

💔 Test failed - checks-actions

[tgross35]

Forgot to rename a function, could use another try please

[Noratrieb]

try harder
@bors try

[bors]

⌛ Trying commit 4c51d7f with merge c6be106...

[Kobzol]

These spurious failures are really starting to compound...

@bors retry

[bors]

⌛ Testing commit 42ac903 with merge 678c1bd...

[rust-log-analyzer]

The job x86_64-msvc-ext failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

[RUSTC-TIMING] miri test:false 65.058
[RUSTC-TIMING] miri test:false 3.061
    Finished `release` profile [optimized] target(s) in 0.41s
thread 'main' panicked at src/lib.rs:1668:17:
failed to copy `C:\a\rust\rust\build\x86_64-pc-windows-msvc\stage1-tools\x86_64-pc-windows-msvc\release\miri.exe` to `C:\a\rust\rust\build\x86_64-pc-windows-msvc\stage1-tools-bin\miri.exe`: The process cannot access the file because it is being used by another process. (os error 32)
##[endgroup]
Build completed unsuccessfully in 0:00:10
  local time: Tue, Jul  2, 2024  3:17:39 PM
  network time: Tue, 02 Jul 2024 15:17:40 GMT

[bors]

💔 Test failed - checks-actions

[dtolnay]

@bors retry

[bors]

⌛ Testing commit 42ac903 with merge bcf38b5...

[bors]

☀️ Test successful - checks-actions
Approved by: Kobzol
Pushing bcf38b5 to master...

[rust-timer]

Finished benchmarking commit (bcf38b5): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

This benchmark run did not return any relevant results for this metric.

Max RSS (memory usage)

This benchmark run did not return any relevant results for this metric.

Cycles

Results (secondary 4.7%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	4.7%	[4.4%, 5.0%]	8
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 695.108s -> 696.686s (0.23%)
Artifact size: 327.55 MiB -> 327.51 MiB (-0.01%)

[cuviper]

I'm unilaterally backporting this to unblock #127377.

@rustbot label +beta-accepted

[tgross35]

I'm surprised that it is still down.

@he32 (since you're listed as the NetBSD maintainer) any clue if anyone from NetBSD are aware of this? Summary: the CDN for 9.0 and 9.1 have been 404ing for a few days (e.g. https://cdn.netbsd.org/pub/NetBSD/NetBSD-9.0/source/sets/). Other versions correctly redirected to the archive URL for a while while the CDN was down, and now seem to be back up at the CDN (e.g. https://cdn.netbsd.org/pub/NetBSD/NetBSD-9.2/source/sets/) - but 9.0 and 9.1 CDN links are just dead.

This isn't currently a problem for Rust since we switched to the archive URL (this PR), but presumably we aren't the only ones to experience some fallout.