Fixed "user is not managed" error when accessing ElastiCache and MemoryDB

[greedy52]

Fixes #30297

$ tsh db connect --db-user teleport-readonly steve-cluster-redis
localhost:55328> acl whoami
(error) ERR Teleport: Could not connect to database: failed to get password for teleport-readonly: database user teleport-readonly is not managed. Make sure that IAM auth is enabled for ElastiCache user "teleport-readonly" and Teleport database agent's IAM policy has "elasticache:Connect" permissions (note that IAM changes may take a few minutes to propagate): { "Version": "2012-10-17", "Statement": [ { "Effect": "Allow", "Action": "elasticache:Connect", "Resource": [ "arn:aws:elasticache:ca-central-1:<account>:replicationgroup:steve-cluster-redis", "arn:aws:elasticache:ca-central-1:<account>:user:*" ] } ] }
localhost:55328> 

We had a few recent changes (#29215, #29864) to make copies of database objects in Database Access. This broke ElastiCache/MemoryDB managed users as the internal tracking was relying on comparing database objects by pointer instead of by name.

Changed the internal tracking to cache by database.GetName() and adds some extra handling when a database name is reused for a new database.

[public-teleport-github-review-bot]

@greedy52 See the table below for backport results.

Branch	Result
branch/v11	Create PR
branch/v12	Create PR
branch/v13	Create PR