| 12 Feb 2025 |
 Atemu | There's a doc on local mirroring | 08:06:54 |
| Atemu | I don't remember the exact env var name | 08:07:08 |
| Atemu | You can just use a remote mirror instead | 08:07:32 |
 waltmck | found it | 08:07:32 |
| waltmck | okay, I followed the instructions to set a local git mirror at /cache/mirror and but got the following error:
subprocess.CalledProcessError: Command '['git', 'ls-remote', '/cache/mirror/platform/external/android-key-attestation']' returned non-zero exit status 128.
This also fails from the command line, sine it is the wrong location: the correct location is /cache/mirror/external/android-key-attestation. I tried moving /cache/mirror to /cache/mirror/platform so that it would find the correct location, but then it fails with
subprocess.CalledProcessError: Command '['git', 'ls-remote', '/cache/mirror/device/generic/goldfish-opengl']' returned non-zero exit status 128
It seems like it is looking in the wrong location for external/*, any idea what the reason for this could be?
| 21:09:36 |
| waltmck | BTW this error happens even before I add my device to supported_devices, so I don't think it's particular to that | 21:11:15 |
| waltmck | * okay, I followed the instructions to set a local git mirror at /cache/mirror and but got the following error:
subprocess.CalledProcessError: Command '['git', 'ls-remote', '/cache/mirror/platform/external/android-key-attestation']' returned non-zero exit status 128.
This also fails from the command line, since it is the wrong location: the correct location is /cache/mirror/external/android-key-attestation. I tried moving /cache/mirror to /cache/mirror/platform so that it would find the correct location, but then it fails with
subprocess.CalledProcessError: Command '['git', 'ls-remote', '/cache/mirror/device/generic/goldfish-opengl']' returned non-zero exit status 128
It seems like it is looking in the wrong location for external/*, any idea what the reason for this could be?
| 21:12:18 |
| Atemu | Wait, you actually set up a mirror? | 22:04:09 |
| Atemu | As in the full 1.4TiB? | 22:04:26 |
| Atemu | What I meant was to use the github mirror | 22:04:49 |
| Atemu | Though thinking about it again, it has a different structure because it can't do sub-paths, only owner/repo | 22:05:49 |
| waltmck | It's only 126GB. I followed these instructions | 22:28:21 |
| Atemu | That's just a local checkout | 22:29:04 |
| Atemu | The full mirror has separate instructions and is enormous | 22:29:29 |
| waltmck | your documentation calls it a local mirror, but sure I see | 22:29:40 |
| Atemu | Oh neat Google removed the instructions on how to set up your own mirror | 22:31:03 |
| waltmck | so you're saying issue would apply to a github mirror as well? | 22:31:12 |
| Atemu | That's always great to see | 22:31:13 |
| Atemu | It'd be a different kind of issue | 22:31:32 |
| Atemu | Though perhaps it could be worked around in the mirror handling code? | 22:31:55 |
| Atemu | Unless you have a full mirror, waiting for google to like you again and then not doing it quite as hard is your best bet | 22:32:46 |
| Atemu | An exponential backoff retry mechanism would be great but I wouldn't put too much effort into the updaters, given that a replacement is in the works | 22:34:05 |
| Atemu | * | 22:34:20 |
| waltmck | Have you been blacklisted by google's servers before? how long did you have to wait? | 22:35:10 |
| Atemu | Yes and only a few hours I think | 22:35:32 |
| waltmck | I tried after 8 hours again with 1 thread and it still didn't work | 22:35:46 |
| Atemu | You should be good to try right about now | 22:35:46 |
| Atemu | Hm | 22:35:55 |
| waltmck | sure, I'll try again now. It's been about 4-5 hours since my last try | 22:36:01 |
| Atemu | Well as I said, your guess is as good as mine | 22:36:15 |
