|
|
|
|
|
|
by kmac_
1310 days ago
|
|
|
1. 1000 RPCs is nothing surprising in any large solution. There are layers and layers of specialized services that are called parallelly and asynchronously. Can a "local" service handle everything by itself? Check internal datacenter letencies. Most of the calls are culled by caches anyway. Also those calls are made with Thrift/Protocol buffers/gRPC. Such design allows scaling, resiliency and prevents breaking other services by circuit breaking. I could go on and on. 2. How do you know that Twitter employees were disengaged? |
|
|
Some are posting "kiss my ass Elon" on Twitter. Other's tweets mocking the code review and "hardcore" engineer stuff. I'm sure you're going to say "But that doesn't mean all of them" but Musk hasn't fired all of them.