I've been working on my project for 6 months, I've sacrificed everything for the project and I feel like I may have done a big mistake depending on GAE. And I am sure I spent 5 months out of 6 months on GAE specific things.
Let's say I have an application that depends on ad income. And assume every request is 1 second and Task Scheduler is perfect! So I pay $0.05 to an instance every hour, which has 3600 seconds inside it. So lets average that number to 3600 requests and lets say there are 1800 page views. (Assuming things are done with Ajax). So the cost for 1000 page views are: $0.05/1800*1000=0.027$. Assuming everything works perfect, and not counting background tasks, although mine are huge, I need to get at least 0.027$ ecpm. For example for Turkish traffic sometimes ecpms can drop below 0.1$'s, and I am sure there are countries out there with significantly lower ecpms and our cost traffics were optimal. To sum up, It seems if I use GAE, I will always have the risk that the costs will be higher than the income ... I have applications on Dedicated Servers, usually a server is nearly idle, the load is 0.5 out of 8, sometimes 2-3, and I am sure I don't utilize them to %20 maybe, but still my cost ratio is %10 ! If I could utilize a server to a maxomum level that could be %2 ! And on the best case it seems this rate will be %25 on gae assuming everything is perfect .... This pricing and advertising under the page: http://www.google.com/enterprise/appengine/appengine_pricing.html made me think gae only wants client like Best Buy etc, big companies who have much higher income rates from web products ... Other than these cost problems, I am using Python and I am very worried since Go came out, which seems to be run multi-threaded in future, Java is also multi-threaded, and as Python users we will only have 1 instance / 1 request. So is Python GAE feasible at this point? it doesn't seem that way? On May 11, 3:40 am, Albert <[email protected]> wrote: > I just checked the new proposed pricing > here...http://www.google.com/enterprise/appengine/appengine_pricing.html > > I'm confused why all the items below "Channel API" in the API Pricing > models have check marks instead of a price per unit. What does that > mean? > > And when they say, "Frontend Instances", does that include instances > handling task queues and crons? > > Thanks! > > Albert > > On May 11, 8:24 am, Ugorji <[email protected]> wrote: > > > > > > > > > eeIt's actually stated in the > > blog:http://blog.golang.org/2011/05/go-and-google-app-engine.html > > > Also, although goroutines and channels are present, when a Go app runs on > > App Engine only one thread is run in a given instance. That is, *all > > goroutines run in a single operating system thread, so there is no CPU > > parallelism* available for a given client request. We expect this > > restriction will be lifted at some point. > > > So you can still use go routines, channels, etc - but we're back to like the > > days of green threads in java where the runtime multiplexes them on a single > > thread (which is fine). However, we don't get concurrent web requests on the > > same instance (which is not fine). Consequently, right now, Java Runtime > > seems to have a pretty significant advantage over the others (even over GO > > which has concurrency as some of its major advantages). And with instance > > pricing, it seems like it directly affects cost. -- 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.
