Autoscale config has not updated hard limit

Not sure if this is a temporary error or not as 60 was the previous setting, but it has been 40 minutes which seems on the long side. It should be a hardlimit of 160 as per the config uploaded.

This instance was deployed a while ago.

d37d7926        app     105     mia     run     running 1 total, 1 passing      0               43m20s ago
2022-08-08T18:40:46Z proxy[d37d7926] mia [warn]Instance reached connections hard limit of 60
2022-08-08T18:40:46Z proxy[d37d7926] mia [warn]Instance reached connections hard limit of 60
2022-08-08T18:47:32Z proxy[d37d7926] mia [warn]Instance reached connections hard limit of 60
$ fly autoscale show
     Scale Mode: Balanced
      Min Count: 3
      Max Count: 6
$ fly config display 
{
    "env": {},
    "experimental": {
        "allowed_public_ports": [],
        "cmd": [],
        "entrypoint": [],
        "exec": []
    },
    "kill_signal": "SIGINT",
    "kill_timeout": 5,
    "processes": [],
    "services": [
        {
            "concurrency": {
                "hard_limit": 160,
                "soft_limit": 100,
                "type": "connections"
            },
            "http_checks": [],
            "internal_port": 4000,
            "ports": [
                {
                    "force_https": true,
                    "handlers": [
                        "http"
                    ],
                    "port": 80
                },
                {
                    "handlers": [
                        "tls",
                        "http"
                    ],
                    "port": 443
                }
            ],
            "processes": [],
            "protocol": "tcp",
            "script_checks": [],
            "tcp_checks": [
                {
                    "grace_period": "1s",
                    "interval": "15s",
                    "restart_limit": 25,
                    "timeout": "2s"
                }
            ]
        }
    ]
}


Had to re-deploy with a new docker image to fix which re-created all the vms.