Segmentation fault on db pull

[svickers]

Describe the bug
I've had no luck getting supabase set up for local development with an existing project. All the containers start fine but I end up with a blank file migrations/20241229224617_remote_schema.sql file and a segmentation fault error. I've tried db reset multiple times as well but with no luck.

To Reproduce
Steps to reproduce the behavior:

1. npx supabase init
2. npx supabase login
3. npx supabase link
4. npx supabase start
5. npx supabase db pull --linked

Expected behavior
New file in /migrations that contains the init script for the database.

Screenshots

System information

• Ticket ID: 170026373437488bb89e4a8ba115ea6b
• Version of OS: Windows 11
• Version of CLI: 2.1.1
• Version of Docker: 27.4.0
• Versions of services:

SERVICE IMAGE	LOCAL	LINKED
supabase/postgres	15.8.1.020	15.8.1.020
supabase/gotrue	v2.167.0	v2.167.0
postgrest/postgrest	v12.2.3	v12.2.3
supabase/realtime	v2.33.58	-
supabase/storage-api	v1.14.4	v1.14.4
supabase/edge-runtime	v1.65.4	-
supabase/studio	20241202-71e5240	-
supabase/postgres-meta	v0.84.2	-
supabase/logflare	1.4.0	-
supabase/supavisor	1.1.56	-

[sweatybridge]

This is a bit strange. Could you try the following command to help rule out potential issues with the postgres image

echo '15.6.1.143' > supabase/.temp/postgres-version
npx supabase db pull --linked

[svickers]

@sweatybridge Same issue still unfortunately.

[auskim]

Seeing this same segfault message, on CLI version 2.2.1, macOS ARM binary. I also rolled back to 2.0.0 and manually ran the pg_dump commands against our Supabase remote DB, all with the same result. I tried the pg_dump command with every --schema and --table we have, and those do actually all work.

I didn't see this problem earlier today. supabase db pull worked, but I reverted the migration history on the remote and cleared all local migration files to try and do a 'clean pull' and I believe this issue started immediately afterward.

[auskim]

Seems to be resolved by pushing a empty new migration to the remote DB from a clean local instance (no existing schemas/tables/migrations) via supabase db push, which works for some reason, and then trying supabase db pull. It's not really clear to me why

[sweatybridge]

For future reference, another alternative you can try is to dump the schema without docker

supabase db dump --dry-run > dump.sh

This creates a bash script that you can then run to fetch the remote schema.