Hi guys, we are having problems with one of our D1 databases - we can't seem to be able to reach it

Hi guys, we are having problems with one of our D1 databases - we can't seem to be able to reach it using either the web portal or from a worker.

FarhianOP•9/19/23, 1:54 PM

All the other databases are ok - only one seems to be acting up.

FFarhian All the other databases are ok - only one seems to be acting up.

James•9/19/23, 2:01 PM

What error are you seeing? Internal error: failed to get DB size and num_tablesInternal error: failed to get DB size and num_tables?

FarhianOP•9/19/23, 2:03 PM

"Failed to reach database. Please try again later."

James•9/19/23, 2:04 PM

@elithrar Seems like this error is growing - ~4 reports now I think

FarhianOP•9/19/23, 2:05 PM

(and I know that we 'should' not be using D1 for Prod - but we are :))

FFarhian Hi guys, we are having problems with one of our D1 databases - we can't seem to ...

Matt Silverlock•9/19/23, 2:20 PM

Please share the databaseId.

FarhianOP•9/19/23, 2:20 PM

sec

FarhianOP•9/19/23, 2:21 PM

@elithrar I've sent you the DB id in private

FFarhian @elithrar I've sent you the DB id in private

Matt Silverlock•9/19/23, 2:22 PM

Thanks. DB IDs can be public — no one can do anything with your DB ID.

FarhianOP•9/19/23, 2:22 PM

oh ok

Matt Silverlock•9/19/23, 2:22 PM

(I have it now, but for future)

FarhianOP•9/19/23, 2:23 PM

sounds good, thanks!

FarhianOP•9/19/23, 2:38 PM

@colexemi yes, the error from the networking tab is: "errors": [
{
"code": 7500,
"message": "Internal error: failed to get DB size and num_tables"
}
],

Matt Silverlock•9/19/23, 3:25 PM

@farhian @colexemi @Jeff -

we're looking to roll out a patch this afternoon that should address this (we believe all of your DBs are affected by this edge case).

FarhianOP•9/19/23, 3:25 PM

Thank you @elithrar !

dave•9/19/23, 3:27 PM

what happens if I delete a D1 database that is currently bound to two Workers?

dave•9/19/23, 3:27 PM

for context, I'm deleting a database that's in alpha, to recreate it as beta

FarhianOP•9/19/23, 3:30 PM

I think that the process should be (not an expert - just my thoughts)

Create the beta DB and migrate the data from the alpha
Update your worker's toml file in order to update the bindings to point the new DB
re-deploy your workers using the new bindings
delete the alpha DB

dave•9/19/23, 3:38 PM

oh, that's smarter

dave•9/19/23, 3:38 PM

I was gonna delete first lol

FarhianOP•9/19/23, 3:38 PM

Cyb3r-Jak3•9/19/23, 3:38 PM

I don’t think it will let you delete a binding in use

dave•9/19/23, 3:42 PM

hmm, can I have a duplicate name database though?

FarhianOP•9/19/23, 3:45 PM

I do not think so... what I did (which is probably NOT a good practice) was to do a backup first (you need it anyway for the migration), delete it (it did allow me to delete even if I had a worker using it) - like you I wanted to use the same name, went through the beta migration, updated the toml binding and used the new ID and then redployed the worker. In my case, it was OK to have an outage

FarhianOP•9/19/23, 3:52 PM

or the same one