Error: failed moving app: This feature is not available for Postgres apps at this time

1 
? Move app rotator-api-db? Yes
Error: failed moving app: This feature is not available for Postgres apps at this time

I just moved a PG app like this from a different org to a new org - So I know this is possible at this time, how can we get this done?

Thanks!

2

We have tried creating a new app via snapshots, but these new apps will also not move to a new org.

There is lots of topics with similar issues and no real solutions, can someone share a solution to move a PG app to a new org please?

When trying to fork in the same org in the hopes that the new app will actually migrate we get these errors:

failed post-init: failed to enable repmgr: failed to create repmgr database: ERROR: template database "template1" has a collation version mismatch (SQLSTATE XX000). Retrying...

Edit:

I have tried absolutely everything possible to create a clone of the current PG app by creating from snapshot or trying to create from a fork, nothing will result in a healthy new app. Nothing works.

We have now tried to use remote backups - this also fails with no helpful error messagesā€¦

fly pg backups create --name move-to-scratch --app rotator-api-db
Exit code: 1
Performing backup...
failed to create backup: signal: killed
Error: failed to create backup: signal: killed

This suggestion / solution (Moving postgres app to another org - #2 by flyandre) makes no sense, you canā€™t access snapshots across orgsā€¦?

3

Moving across orgs is a little tricky as they involve IP changes. I would recommend creating a new Postgres cluster within the target organization and using fly pg import to import your data.

4

Thanks @shaun - yes I finally resorted to this, I had a different PG app that moved flawlessly, not sure what the difference was. There was issues outside moving orgs here by the way, snapshot and forking had tons of issues within the same org in the hopes that a new flesh version of the app would move like my other PG app, but I couldnā€™t even get a same org clone going.

5

More issues trying to redeploy apps after attaching new DATABASE_URL:

app hanging on waiting for healthy allocations:

[PM01] machines API returned an error: "machine ID 82d335a767e348 lease currently held by dan.wetherald@gmail.com, expires at 2024-07-17T22:15:04Z"

More errors:

WARN failed to acquire lease: failed to get lease on VM 28606dbee09238: machine ID 28606dbee09238 lease currently held by dan.wetherald@gmail.com, expires at 2024-07-17T22:22:01Z

Error: error acquiring leases on all machines
6

Sometimes interrupts can leave leases around for a minute or two.

Assuming you donā€™t any active operations on-going, you can manually clear the leases with:

fly machines leases clear <machine-id>
7

Looks like your source App was running an older versionā€¦ Thereā€™s a known issue when migrating from < v0.0.43 to >= v0.0.43 that can lead to collation issues.

Itā€™s annoying, but can be manually addressed by following the instructions here:

Alternatively, you can perform pgdump/pgrestore which is what the fly pg import process does behind the scenes.

8

Got it, yes we went with fly pg import with the tutorial above with private ipv6 address across orgs.

1 Like
9

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.