@VInuth, if you have 5-10 requests per second and as many urls, it might be easy to calculate cost. But large applications with 100s of instances with as many requests per second and urls getting loaded, its hard to say where you will end up.
I agree on latency part. For instance, the urlfetch latency is really dependent on the url being called. Besides that, such calls dont necessarily mean high memory usage or CPU usage. As Waleed said earlier, instance should be optimized to process requests when urlfetch is waiting for the response. Charging for such instances is very much an Amazon way of thinking, which was the only reason we picked AppEngine over AWS. On the whole, AppEngine team is looking for their hockey stick growth in revenues and pageviews. Not sure if their internal metric of growth was met, but going by this huge change they are making, they might be lagging. And to fix that, we have this drastic change to deal with. In my opinion, simpler change would have been supporting PHP and get tons and tons of new developers for free. For all the AppEngine team bias for PHP, PHP has the ability and volume to make and break a product like AppEngine. As for success of AWS, i think it might be more to do with the fact that you can grab a server and put PHP on it and get going, which besides Amazon, each and every hosting company has realized over the years except for Google. On Jun 28, 8:02 am, Vinuth Madinur <vinuth.madi...@gmail.com> wrote: > I think it is fairly easy to calculate what your costs will be based purely > on the latency of your requests and the number of requests you are getting. > The new scheduler will anyway not affect this and based on scheduler knobs + > front end instance resources, you can somewhat predict the no.of instances > you might require. > > My gripe is, the new pricing brings latency into focus, while the developers > have nothing but their app code to optimize it. The responsibility for > latency is both on the application as well as underlying infrastructure. > > > > > > > > On Tue, Jun 28, 2011 at 5:04 PM, Albert <albertpa...@gmail.com> wrote: > > I'm holding 2 workshops for Google AppEngine next month. I used to > > look forward to encourage developers to use GAE as a development > > platform, but due to this new pricing changes, I'm not that excited > > anymore. It's not because I don't like GAE anymore. I believe that the > > GAE infrastructure offers a lot of value as an infrastructure. > > However, I really think that the pricing changes announcement was > > announced to early. The early announcement has caused a lot of > > confusion over a long period of time even until now. The most basic > > question that needs to be answered is "How much will our apps cost to > > run now?" Maybe sample apps and their corresponding prices will help > > us visualize if we should really worry about the upcoming changes. > > > Something like... > > > ------ > > Sample App 1 (datastore intensive app) > > > Average Response Time: 200ms > > > Average # of users per day: 100K > > > Cost / day: ??? > > > ------ > > > Sample App 2 (Compute intensive app) > > > Average Response Time: 800ms > > > Average # of users per day: 100K > > > Cost / day: ??? > > > ------ > > > and so on... > > > ------ > > > At least I can get a kind of "official" idea of the costs of running > > apps under the new pricing scheme. > > > Thanks and enjoy! > > > Albert > > > On Jun 28, 1:23 am, vivpuri <v...@vivekpuri.com> wrote: > > > There is one more option - Application just has to shutdown since > > > there is no money to pay. If a bill of $2k per month suddenly becomes > > > $10k per month, there are not many who can pay that. Just for > > > comparison sake, in the US mortgage crisis, at the end of 5ARM a > > > sudden increase in 1% of interest rate(effectively $200-$500 dollars/ > > > month, depending on total amount) lead home owners to default and file > > > bankruptcy. > > > > On Jun 27, 12:50 pm, Branko Vukelic <bg.bra...@gmail.com> wrote: > > > > > On Mon, Jun 27, 2011 at 6:44 PM, vivpuri <v...@vivekpuri.com> wrote: > > > > > In my opinion, this pricing change has ability to destroy AppEngine. > > > > > I believe you are underestimating the amount of bad experience users > > > > that are locked into the platform will put up with before they switch. > > > > I'm sure Google could have done some PR before this change, and > > > > prepped the users for the change, but I'm also sure most users will > > > > just bite the bullet and pay up. Sure, some will leave, but I bet most > > > > will stick to GAE. > > > > > If you coded your application without any layer of abstraction, and > > > > your code is highly optimized for running on GAE, it costs more to > > > > move away from it, than to sustain the increased fees until you can > > > > monetize your application. Of course, if the application wasn't meant > > > > to be used for business, that's different. If it's a hobby, you can > > > > 'afford' to move to another platform. But for a business that is > > > > planning on monetizing, moving is just as expensive as staying. > > > > > -- > > > > Branko Vukelić > > > > bra...@herdhound.com > > > > > Lead Developer > > > > Herd Hound (tm) - Travel that doesn't bitewww.herdhound.com > > > -- > > You received this message because you are subscribed to the Google Groups > > "Google App Engine" group. > > To post to this group, send email to google-appengine@googlegroups.com. > > To unsubscribe from this group, send email to > > google-appengine+unsubscr...@googlegroups.com. > > For more options, visit this group at > >http://groups.google.com/group/google-appengine?hl=en. -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com. To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
