Thank you! Every time a story like this comes up, people seem to wanna pretend managing your own hardware is all sunshine and rainbows. Especially if you want global scale or as little down-time as possible, cloud provider’s your best bet, albeit one where you have less control than you would with your own servers.
Opinion: You should be building on top of open source platforms and tools (Docker, Kubernetes if you need it…granted I’m not an expert in this area) to mitigate some of the vendor-lockin, and take a multi-cloud approach. If you’re mainly hosting on GCP for example, host smaller deployments on AWS, Azure, Cloudflare, or something else as a contingency…eventuality you can also add or just move to your own servers relatively painlessly. Also AGGRESSIVELY backup up your database in multiple places.
Thank you! Every time a story like this comes up, people seem to wanna pretend managing your own hardware is all sunshine and rainbows. Especially if you want global scale or as little down-time as possible, cloud provider’s your best bet, albeit one where you have less control than you would with your own servers.
Opinion: You should be building on top of open source platforms and tools (Docker, Kubernetes if you need it…granted I’m not an expert in this area) to mitigate some of the vendor-lockin, and take a multi-cloud approach. If you’re mainly hosting on GCP for example, host smaller deployments on AWS, Azure, Cloudflare, or something else as a contingency…eventuality you can also add or just move to your own servers relatively painlessly. Also AGGRESSIVELY backup up your database in multiple places.