If I run the sandbox against a PostgreSQL HA configuration, it can happen that the database connection is dropped during failover or scaling operations. Does the sandbox gracefully recover from such a situation and will try to reconnect? Or do I need to restart the process in such a situation?
2 Likes
Right now there is a built-in capability to recover from transient errors. What you would experience when using the Ledger API are transient INTERNAL errors.
However, I would highly encourage operators to make high availability a deployment concern as well.
1 Like