Nothing in upaas said that a deploy keeps only an app's volume mounts, and webhooker lost its database at one.
When an app has no volume mounts, its page's Volume Mounts section says: "This app has no volume mounts, so the files it writes are lost whenever a deploy or rollback replaces its container; a restart of the container keeps them."
Each deploy of such an app writes the same sentence, one constant in the deploy package, into its log once, after the webhook payload and before the clone.
The README's Volume mounts section says a deploy or rollback replaces the app's container with a new one, which keeps only the files the app wrote to its volume mounts.
Restart stops and starts the same container; deploy and rollback remove it and create a new one.
Disclosures:
Judgement call: the app page has no Restart button (only a restart route nothing links to), so the sentence says "a restart of the container"; Docker's own restarts after a crash or reboot keep the files too.
Deviation: "whenever a deploy or rollback replaces its container", not "at every deploy", because a deploy whose build fails leaves the old container and its files in place.
Judgement call: a rollback's log does not carry the sentence; the issue asks for deploy logs only.
If reading the volume mounts fails, the log leaves the sentence out; the deploy then fails with that error when it creates the container.
Model: opus-5-5
Nothing in upaas said that a deploy keeps only an app's volume mounts, and webhooker lost its database at one.
- When an app has no volume mounts, its page's Volume Mounts section says: "This app has no volume mounts, so the files it writes are lost whenever a deploy or rollback replaces its container; a restart of the container keeps them."
- Each deploy of such an app writes the same sentence, one constant in the deploy package, into its log once, after the webhook payload and before the clone.
- The README's Volume mounts section says a deploy or rollback replaces the app's container with a new one, which keeps only the files the app wrote to its volume mounts.
Restart stops and starts the same container; deploy and rollback remove it and create a new one.
Disclosures:
- Judgement call: the app page has no Restart button (only a restart route nothing links to), so the sentence says "a restart of the container"; Docker's own restarts after a crash or reboot keep the files too.
- Deviation: "whenever a deploy or rollback replaces its container", not "at every deploy", because a deploy whose build fails leaves the old container and its files in place.
- Judgement call: a rollback's log does not carry the sentence; the issue asks for deploy logs only.
- If reading the volume mounts fails, the log leaves the sentence out; the deploy then fails with that error when it creates the container.
Model: opus-5-5
An app with no volume mounts now shows, in its Volume Mounts section,
that the files it writes are lost whenever a deploy or rollback replaces
its container and that a restart keeps them. Each deploy of such an app
writes the same sentence into its log, after the webhook payload and
before the clone. The README says in one sentence that a deploy or
rollback starts a new container that keeps only the files in the app's
volume mounts.
Model: opus-5-5
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Nothing in upaas said that a deploy keeps only an app's volume mounts, and webhooker lost its database at one.
Restart stops and starts the same container; deploy and rollback remove it and create a new one.
Disclosures:
Model: opus-5-5
Review passed.
Model: opus-5-5
80a15adfa3to396d3ef064Review passed.
Model: opus-5-5