App Engine is finished, here's why

[Kenneth]

What has always been the biggest concern about app engine?  Lock-in.  You're at the mercy of Google.  Sure there's typhoon ae etc... but really those are not alternatives.  

What does Google go ahead and do?  They do exactly what their critics said they would do and what us GAE adopters hoped like hell they would never do, screw us over.

App Engine is finished not because we're all going to move off to EC2, but because people who are considering using app engine will see exactly what has gone on here with the pricing, think about the lock-in argument against GAE, and decide not to use GAE.  There will be a drop off in new apps, and eventually Google is going to see GAE isn't really panning out and pull the 3 year plug.

http://en.wikipedia.org/wiki/K%C3%BCbler-Ross_model, I'm at stage 4.

[Raymond C.]

I see it's end too:

- those who are leaving will absolutely never look back again

- those who left will never recommend anyone to use GAE again or even recommend to stay away from Google's product because of the awful experience

- if GAE havent took off in the past 3 years, I dont see why it will after this change

-> a platform with not enough user will just being faded out naturally

GAE has never been a supervisor product in my opinion.  We bet on Google, hoped for the best, and now Google intend to destroy everything we've built in the past three years.

[Martin Waller]

Hello,

Everyone seems to believe that GAE is finished. I'd like to hear where people are planning to move their applications too, both the Python users and the Java users?

Martin

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.
To view this discussion on the web visit https://groups.google.com/d/msg/google-appengine/-/Ri0fGinW-x8J.
To post to this group, send email to google-a...@googlegroups.com.
To unsubscribe from this group, send email to google-appengi...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.

[Daniel Florey]

If it is true that my app billing will increase from ~5$ per day to ~70$ per day as stated in the new billing preview, I'll go back 10 years in time and will host our apps on dedicated root servers.

For the same money that I'll have to spend on appengine I'll get 5 dedicated hexacore servers with 16 gig of memory and 4TB of storage each.

So this will be the route I'm forced to go unless Google is waking up and recalculates the new pricing scheme.

Of course this means to trash years of work and start from scatch regarding the datastore (moving from datastore -> sql) and to deal with the hassle of apache load balancing etc.

[Strom]

My app gets about 70x price increase, but I'm not going to go anywhere
else.
Even with such a price increase, GAE is better than the alternatives.
Sure that money would buy me quite a few dedicated servers.
But it does not buy me a team of people who will fix issues at 4AM in
the morning.

[Caglar Cataloglu]

Hi guys,

I am using GAE free edition and after reading these messages i am a bit disappointed now.

2011/9/1 Strom <xxs...@gmail.com>

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.

[Gopal Patel]

I remember some talks that, google is going for instance-hour for better predictability for resource allocation and revenue.

here are my free suggestions, that can make app engine looks bit good....

-- keep free quota but keep small amount for an app. i.e. 5$ per year per app. ( that way, you will get few tens millions or more ) .( perfectly predictable ). in return allow unlimited app and dynamically creating app through API. ( some of  per customer separate app will see this as big advantage . )

-- match competition. aws and azure dropped incoming bandwidth cost.

-- i think cpu going faster, memory and bandwidth going cheaper, give us continuous price reduction ( the older an app, cheaper the pricing, we are not junk that you can use to test your free platform and once it get mature, ask us to paid more.. ) and you are using us for creating platform of the future, so return it by giving some price reduction for older apps.

-- make ssl on https ( seriously, what is taking so long, even if it is rocket science, you guys got smartest mind to do it ) free and also add mysql and make it free( of course regular charge for data, bandwidth, cpu applys...no separate billing for instance..).

-- what is done inside your api and your back end datastore servers is not our concern. ( it is designed by you and if it uses separate query for list of key  and fetching that entities by looping, its not my concern. combined them all in to one package. and add it in to datastore storage price. ).

-- .....and I have read optimization article...one thing I would like to point out is, since it is confirm that a instance is not going to die in 15 minutes, there is a big chance that instance memory is going to be a killer feature for latency reduction. ( cache invalidation is problem, so give us "broadcast to all instance api". please. )

after all of above, I will not argue much about price. promise. and check: I have not said single word about outright price reduction. :P

my 0.02 instance hour. 

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.

To view this discussion on the web visit https://groups.google.com/d/msg/google-appengine/-/Ri0fGinW-x8J.

[Raymond C.]

There are quite some PAAS/IAAS options out there which don't require you to have "a team of people who will fix issues at 4AM in 
the morning"

[Strom]

What PAAS options do you have in mind? Azure? If we're comparing GAE
to Azure, then we shouldn't take the price of a single Azure instance
as GAE apps have high replication by default.

[Raymond C.]

Heroku could be the most famous one.  Its seems to have good fame on its free tier for small and quick web apps.

[Ravi Sharma]

I 100% agree with everyone here but for me main concern was how Google doing all this business......
First they gave us one crapy plateform for free, we worked hard on it, we became free tester for them...we helped google to make it better....we spent 100s of hours in doing that...... and once its in condition that some can use it with much more confidence they are charging whatever they think is appropraite......
the first and main reason for me to try GAE was cost and now its not there...we can always compare technolgy and services with other platform...but if GAE doesnt suit my Costing/budgeting i wont continue on it, it may be 1000% better then others.....
i m just saying its kind of business stratgey(what i have seen in movies)...first give free drugs to people and once they are habitual charge them wahetver u want :).

My 0.003 cpu instance of emotions.....

Thanks
Ravi.

On Thu, Sep 1, 2011 at 12:07 PM, Raymond C. <wind...@gmail.com> wrote:

Heroku could be the most famous one.  Its seems to have good fame on its free tier for small and quick web apps.

--

You received this message because you are subscribed to the Google Groups "Google App Engine" group.

To view this discussion on the web visit https://groups.google.com/d/msg/google-appengine/-/XjUEtZ_tGG0J.

[Kenneth]

My concern isn't people moving, there may be some of those but I'm not one of them.  Any cost savings I might hypothetically get from moving to another platform would be wiped out by several orders of magnitude on the cost of the port.  I'm also not at all convinced another platform would actually be cheaper looking at the whole picture.  I'm just depressed that the margins on my business have gotten that much thinner.

What I am highly concerned about is the message this sends to new customers.  I don't know, maybe that's not valid, it is beta after all.  I think I'm moving to step 5.

[Brandon Thomson]

Kenneth's post describes my feelings pretty well. Changing platforms is completely out of the question for me but I am concerned about the long-term health of GAE. All this bad publicity certainly doesn't do it any good.

I suspect the short 2-week transition period has been chosen because GAE is burning through an obscene amount of money and computing capacity with very little income. The decision makers at Google probably know giving such short notice is going to generate more bad press but thought the situation was severe enough that they should do it anyway.

[Joseph Jude]

don't be evil...oh sorry that was for microsoft.

larger question is: will such 'startegy' decision be on all Google products - gmail, gdocs?

[Anders]

What I think could make customers abandon Google App Engine is having to pay a lot for frontend instances. To use a crude analogy, paying a lot for instances is like paying a lot for SMS messages sent from a cell phone. The quotas for CPU hours, data storage and bandwidth etc are enough to make a huge profit for cloud providers, and the frontend instances are like a cherry on top that is bigger than the cake! Just because other cloud providers are greedy as hell and charge for frontend instances doesn't mean Google must follow that 20th century business strategy.

[vlad]

GAE was fueling its growth by giving a lot of resources for free. Free ride has ended. Now they will have to compete on merits of the platform. Unless they seriously turn to  focus on customers I am worried about GAE.

[Anders]

I guess medium-size applications can run on platforms with much lower prices, such as 4GH: http://www.godaddy.com/hosting/web-hosting.aspx?ci=9009

GAE is probably more scalable in reality though.

[Tapir]

This godaddy plan is for very small apps, not for medium-size apps.
A medium-size app needs a dedicated server.
The cheapest godaddy dedicated server is $70/month:
http://www.godaddy.com/hosting/dedicated-servers.aspx?ci=9014
It is easy to get a 30% discount, so the real price is $50/permonth.
This price is a little cheaper the a always-on GAE instance.
But it has more memory, and the $50 price includes the bandwidth cost,
and the CPU is much much faster than GAE instance.

So if you don't need gae specified features, the godaddy dedicated
server
is more suitable for you.

[GR]

I can't find anywhere that states that it's only for "small" sites... the faq for 4GH states:

How far can I really scale?

There is theoretically no limit. We will continue to add servers and bandwidth to meet demand.

I don't see why you would need a dedicated server that doesn't scale under the 4GH stuff