My “app-db” don’t have any machine now. I can’t deploy app (fly deploy) because database it is not responding. What can I do now?
Hi… What do you see over in the Volumes section?
If you still have one of those, then you can use the volume-forking technique (with an explicit volume ID) or restore from a snapshot.
I have a few. vol_od56vj5ywk94ny30 from year ago is a core volume from which I made some snapshots. I think vol_od56vj5ywk94ny30 is somewhat corrupted because of ARN region problems.
I also have some snapshots
Should I go into forking and then do fly deploy? I will ask chat GPT for help. Any advice is appreciated.
Forking is preferable to restoring from snapshots, since the latter can miss the most recent ~24 hours of writes.
It might take a few tries to find the right volume in this case, though. (This is unusual.)
You can use fly ssh console -a new-db-app-name to get a shell on the new database Machine, after forking, and then check timestamps in various spots under /data (typical location) to verify that they look recent, 
…
1 Like
Thank you
If I have this files, is it mean that my lost images are still there?
➜ scaffold-data git:(main) ✗ fly ssh console -a app-name
Connecting to fdaa:2:6094:a7b:3a:3326:ef65:2… complete
root@d890e3ef1d0278:/rails# ls -la /data/storage
total 96
drwxr-xr-x 24 rails rails 4096 Mar 26 06:16 .
drwxr-xr-x 4 rails rails 4096 Mar 26 06:12 ..
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 07
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 09
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 24
drwxr-xr-x 3 rails rails 4096 Mar 26 06:15 2v
drwxr-xr-x 3 rails rails 4096 Mar 26 06:15 6d
drwxr-xr-x 3 rails rails 4096 Mar 26 06:16 72
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 86
drwxr-xr-x 3 rails rails 4096 Mar 26 06:16 88
drwxr-xr-x 3 rails rails 4096 Mar 26 06:14 8z
drwxr-xr-x 3 rails rails 4096 Mar 26 06:16 9r
drwxr-xr-x 3 rails rails 4096 Mar 26 06:16 9z
drwxr-xr-x 3 rails rails 4096 Mar 26 06:15 bd
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 c0
drwxr-xr-x 3 rails rails 4096 Mar 26 06:16 id
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 k5
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 k8
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 kx
drwxr-xr-x 3 rails rails 4096 Mar 26 06:16 mg
drwxr-xr-x 3 rails rails 4096 Mar 26 06:14 rj
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 tu
drwxr-xr-x 3 rails rails 4096 Mar 26 06:15 tw
drwxr-xr-x 3 rails rails 4096 Mar 26 06:12 zq
root@d890e3ef1d0278:/rails#
Hm… I replied over in the other thread, since this appears to be a continuation of that earlier question (about the Rails Machines’ volumes):
Thank you.
➜ scaffold-data git:(main) fly m list -a xxxxxx-db
No machines are available on this app xxxxxxx-db
I gave up. It seems (according to AI agents like DeepSeek), it is impossible to run a postgres database without machine, and it is impossible to create one.
It is out of my scope to pull images from this storage. I should set a “wire-connection”.
I end up editing my code to use amazon s3 bucket. I was aware and have backup images on my laptop. Two hours of work and did add all photos to parts manually.
Tip for everyone. Go straight with S3 bucket, because with volumes it is hard or impossible to get a clone on different region etc.
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.