Hacker News new | past | comments | ask | show | jobs | submit login

The aws services stack is deep and deeply intertwined. I've always viewed depending on such stacks in production with skepticism and I'd recommend everybody else does that too.

This might come across as tooting our horn a bit. But it's more about sounding a warning to other startups providing SaaS service built on public cloud. My own misgivings about relying on a cloud provider specific stack (both for the reasons of visibility/debuggability as well as for vendor lock-in) meant that PacketZoom services were not affected by this failure at all because we only use them as one of the many providers of raw machines. We use our own techniques to load-balance/failover among multiple cloud providers too (so even if the raw compute/network went away, our service would take a perf hit but not be completely down).

Or you could just run in multiple regions. Using multiple cloud providers limits your ability to take advantage of provider-specific features - why waste time writing your own load balancer when you could use ELB + multiple regions?

"Or you could just run in multiple regions."

Not when the original goal of the very service is to have presence in all geographical regions. If aws us-east is hit, I want the users to transparently failover to a server on east coast (perhaps one hosted by google or softlayer) rather than be directed all the way to us-west or eu.

And as for ELB, one doesn't use ELB for a custom protocol that load-balances/fails-over itself from the client :-)

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact