Google's instance-hour is process-instance-hour whereas AWS's instance- hour is machine-instance-hour. Every process GAE spins up count as one instance-hour while you got the whole machine for a hour in AWS and you can spin up as many processes as the machine instance can handle.
When multiple web requests come in, GAE spins up multiple processes (Python) to serve the requests. All those count as multiple instances. The killer thing is most request handlers are not CPU bound and just idling, waiting for network transfer from browser/ datastore/memcache/etc. All those idling times still count toward the GAE instance-hour. That's why there's a huge increase when switching from pure CPU-hour accounting. In this case GAE's process-instance-hour ($.08/hr) is much more expensive than AWS's machine-instance-hour ($.085/hr or $0.02/hr micro). AWS gives you a whole spec machine which you can cram as many processes out of it. GAE just gives you a process, which idles most of the time for IO bound app but still charges you. Also the spec of the GAE process-instance is pretty vague. You don't know what capability you are buying. On Aug 31, 3:39 pm, Anders <[email protected]> wrote: > I took a look at the Amazon prices just now. It looks quite expensive too. > And the same dubious and opaque cost per instance concept. Both Google and > Amazon can cram in 1000 instances in each physical server! Talk about > hideous profit margin. > > The cost for GAE will be low if the instances is limited to only one. But > what will the performance be then? And what about traffic spikes and/or > traffic growth? -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
