Live data from Hacker News

Is Google App Engine Ready for Prime Time?

sachinrekhi.com

1–10 of 19 posts

Re: Is Google App Engine Ready for Prime Time?

#2
From my own limited experimentation I came to same conclusion as the author. I think that the platform is nice for toy projects because you do get a fair amount of resources for free. However, there are too many limitations for even a serious personal project. This may change once we finally get official pricing.

Re: Is Google App Engine Ready for Prime Time?

#3
Two weeks ago, my client’s datastore indexes became corrupted. The failure was demonstrably AppEngine's. Attempts to manually rebuild indexes failed. The AppEngine team worked for several days to resolve the problem, which we appreciated. But such failures are disheartening when you’re busy signing up paid customers.

I suppose the term “beta” is so overused that it has nearly lost its meaning. Sometimes, however, it still means “warning, ye who enter here.”

--

I've also written some high-level thoughts about the "big three" cloud services on my blog:

http://davepeck.org/2008/11/30/the-clouds-spectrum/

http://davepeck.org/2008/12/03/the-clouds-dark-lining/

Re: Is Google App Engine Ready for Prime Time?

#4

From my own limited experimentation I came to same conclusion as the author. I think that the platform is nice for toy projects because you do get a fair amount of resources for free. However, there are too many limitations for even a serious personal project. This may change once we finally get official pricing.

Pricing's more or less been announced though I guess it's still not "official" yet:

* $0.10 - $0.12 per CPU core-hour

* $0.15 - $0.18 per GB-month of storage

* $0.11 - $0.13 per GB outgoing bandwidth

* $0.09 - $0.11 per GB incoming bandwidth

http://googleappengine.blogspot.com/2008/05/announcing-open-...

It's slightly more expensive than Amazon.

Re: Is Google App Engine Ready for Prime Time?

#5
I recommend staying off of GAE. I was prepared for some bumps, but the failures that I've experienced have, to put it bluntly, made me lose a great deal of confidence in Google in general.

Even if it were more stable, I would still be migrating my company site off of it. The feeling I have now is one of helplessness - my site is going to arbitrarily go down and it'll be Google's fault and there's not a thing I can do about it. It takes me back to the emotions I had in the bad old days when I dealt with Microsoft software in a professional capacity.

Re: Is Google App Engine Ready for Prime Time?

#7
post #4

From my own limited experimentation I came to same conclusion as the author. I think that the platform is nice for toy projects because you do get a fair amount of resources for free. However, there are too many limitations for even a serious personal project. This may change once we finally get official pricing.

Pricing's more or less been announced though I guess it's still not "official" yet: * $0.10 - $0.12 per CPU core-hour * $0.15 - $0.18 per GB-month of storage * $0.11 - $0.13 per GB outgoing bandwidth * $0.09 - $0.11 per GB incoming bandwidth http://googleappengine.blogspot.com/2008/05/announcing-open-... It's slightly more expensive than Amazon.

My problem with the pricing is not that Google hasn't indicated what the pricing will be like, but that if you need more resources there is no other option except hope that Google will be nice to you and up your quotas.

Re: Is Google App Engine Ready for Prime Time?

#9
post #3

Two weeks ago, my client’s datastore indexes became corrupted. The failure was demonstrably AppEngine's. Attempts to manually rebuild indexes failed. The AppEngine team worked for several days to resolve the problem, which we appreciated. But such failures are disheartening when you’re busy signing up paid customers. I suppose the term “beta” is so overused that it has nearly lost its meaning. Sometimes, however, it…

Do you have any pointers to what was wrong and how it was resolved?

Re: Is Google App Engine Ready for Prime Time?

#10
post #4

From my own limited experimentation I came to same conclusion as the author. I think that the platform is nice for toy projects because you do get a fair amount of resources for free. However, there are too many limitations for even a serious personal project. This may change once we finally get official pricing.

Pricing's more or less been announced though I guess it's still not "official" yet: * $0.10 - $0.12 per CPU core-hour * $0.15 - $0.18 per GB-month of storage * $0.11 - $0.13 per GB outgoing bandwidth * $0.09 - $0.11 per GB incoming bandwidth http://googleappengine.blogspot.com/2008/05/announcing-open-... It's slightly more expensive than Amazon.

Not sure if it's more expensive, even if you ignore the free quota. Amazon charges you $0.10 per instance-hour, not cpu-hour like Google (so you only pay while you consume CPU). Also, you should probably compare the storage price not to S3/EBS, but to Simple DB (you still pay for CPU time spent in the datastore, though).
Post reply on HN