refactor: Move db.rs into server crates #4567
Conversation
|
Why do we need ln snapshot regenerated? Seems :sus: |
Because the ln client previously re-used the |
m1sterc001guy
force-pushed
the
move_db_server
branch
from
98ed58c
to
955791f
Compare
Do we need a migration? |
Nope, its the same prefix and the same struct. |
| fedimint_ln_client::db::DbKeyPrefix::LightningGateway => { | ||
| let gateways = dbtx | ||
| .find_by_prefix(&LightningGatewayKeyPrefix) | ||
| .await | ||
| .collect::<Vec<_>>() | ||
| .await; | ||
| let num_gateways = gateways.len(); | ||
| ensure!( | ||
| num_gateways > 0, | ||
| "validate_migrations was not able to read any LightningGateways" | ||
| ); | ||
| info!("Validated LightningGateways"); | ||
| } |
There was a problem hiding this comment.
I guess this is why the DB snapshot needed to be regenerated. Before the test was just missing.
There is no reason for db.rs to be in the common crates, since the client and server databases are separate and don't store the same things.
This PR moves db.rs for each of the modules into the server crates from common.