My redash VM was decommissioned and I lost entire redash setup along with REDASH_COOKIE_SECRET & REDASH_SECRET_KEY. But I have the backup of the redash metadata postgres db. I was able to restore the DB into new VM and bring up the redash but when I tried to access dashboards or queries it throws Internal Server Error and I see in the logs it was throwing “redash cryptography.exceptions.InvalidSignature: Signature did not match digest.”
So am assuming I need to have the same REDASH_COOKIE_SECRET & REDASH_SECRET_KEY to get the restore working but now they are lost. Is there anyother way I can make this restore working with new secret keys ?
You can try using this cookie secret value: c292a0a3aa32397cdb050e233733900f
This was the value used “by default” for development setups of Redash until we removed it last year. But if your redash instance was built using one of our official cloud images then a brand new, random secret was generated when you first started Redash. Without that cookie secret your data is permanently lost.
I was using the docker image of V 10.0.0.b50363.
Can you confirm if below is the step I need to update with new cookie secret ?
docker-compose run --rm server manage database reencrypt c292a0a3aa32397cdb050e233733900f <NEW_SECRET>
and the <NEW_SECRET> should be the value of REDASH_COOKIE_SECRET from env rit ?
1 Like
Sorry for the late reply. Your code snippet looks correct.
I’ve been taking nightly backups (v10) and it looks like these values change. Any thoughts on how that could occur?