0

I am facing issues with the azure private endpoint for managed Postgres. To summarize I have been using Azure Managed Postgres and Private Endpoint already , but one day I accidentally deleted the "managed Postgres" service , Now in my setup I had 1 private endpoint for each Postgres server. Once I deleted the server that endpoint was of no use so I deleted that too.

Now for the past couple of days I have been trying to create a new endpoint but with no success , I was able to create Postgres server with the same name without any issues but endpoint is really a pain. Would appreciate any help. Below is the error I am facing everytime.

{
    "status": "Failed",
    "error": {
        "code": "ResourceDeploymentFailure",
        "message": "The resource operation completed with terminal provisioning state 'Failed'.",
        "details": [
            {
                "code": "ResourceNotFound",
                "message": "The requested resource of type 'Microsoft.DBforPostgreSQL/servers/privateEndpointConnectionProxies' with name 'postgres-qa-ldc-pep.9bc885b4-01af-48b3-8f0c-15749bf07ad1' was not found.",
                "details": []
            }
        ]
    }
}

Now I assume the proxies are all handled by azure in backbone network , how do I fix this ? I already tried and waited multiple days in hope it gets cleaned itself but to no avail.

Appreciate help.

David Makogon
  • 69,407
  • 21
  • 141
  • 189
vgdub
  • 347
  • 5
  • 15
  • thank you @David , I couldn't find that tag. I hope someone help me with some answers. – vgdub Apr 26 '20 at 21:22
  • Are you provisioning the private endpoint on a server name is as the same as a previously dropped server? Please try with a server name that never used before. – x_xiao2_bnu May 01 '20 at 06:58
  • That defeats the purpose , we have set naming convention of the database servers we are provisioning via our IaaC pipeline and the idea was to provision the endpoint just after the server got created, Anyway I think the issue appears to be the time azure takes to clear the old server and it's related endpoint from the backend. I was able to create the same after like 10 hours later or so. Thanks – vgdub May 01 '20 at 11:44

1 Answers1

0

I was able to fix this when I completely removed the parent Postgres server from azure along with the private endpoint for that server and also deleted the Private DNS entry , then I had to wait approximately 10+ hours to get the cleanup from azure done and try the week day morning and it worked at that time. so I assume the wait to clean up time of the old DNS in azure is the issue.

vgdub
  • 347
  • 5
  • 15