@digitalocean yes it got resolved soon, there some issue with backlog of vms in orchestrator/queue I was told, it's all good now. cheers.