Is there something unwell going on at the backend of the snapcraft build environment? I set aside some time to update some snaps. I have had a few failures not of my doing:
error: cannot perform the following tasks:
- Download snap "core18" (2246) from channel "stable" (received an unexpected http response code (401) when trying to download https://storage.snapcraftcontent.com/download-origin/local/CSO04Jhav2yK0uz97cr0ipQRyqg0qQL6_2246.snap?interactive=1&sig=IkNTTzA0SmhhdjJ5SzB1ejk3Y3IwaXBRUnlxZzBxUUw2XzIyNDYuc25hcCI.FGHm0A.Zux8cUWl2Qi94daD2edv4Xpn7tQ)
Install failed
Edit: 30 mins later, it fails again, differently.
From this buildlog:
fatal: the remote end hung up unexpectedly
fatal: early EOF
fatal: index-pack failed
[01/Nov/2021:23:37:33 +0000] "CONNECT github.com:443 HTTP/1.1" 200 1105119471 "-" "git/2.25.1"
Failed to pull source, command 'git clone --recursive https://github.com/Anuken/Mindustry.git /build/mindustry/parts/mindustry/src' exited with code 128.
Build failed
It’s somewhat demotivating to try and update snaps and find they fail to build multiple times in different ways I can’t fix other than by stabbing the “Trigger new build” button.
This mostly seems to be at the layer below Launchpad: DNS, connection timeouts, that sort of thing. (I can’t speak for the GitHub clone error; GitHub could have been flaky at the time, or it could have been a networking issue.) I only have indirect visibility into that sort of thing, but I’ve looked around things like firewall graphs and incident logs and I can’t see anything relevant at that time, and our general rate of build failures doesn’t seem to have been unusually high then either, so I don’t have a lot of leads.
However, one thing that would fairly clearly help here would be telling apt to retry failed connections a few times (it was changed to do that by default recently, but after focal), so I’ve filed a buildd bug to do that.
The snap download failure you linked seems likely to be a store failure of some kind; perhaps a store developer can speak to that.