I'm actually a little surprised it costs so much to run on app engine. Given that you've settled into a more predictable user model in terms of costs and probably growth, why not lower your costs by investing in dedicated server space or possibly reserved instances on AWS (unless there is a app engine equivalent).
Without knowing what your server load looks like, I would imagine you could save a couple hundred dollars a month in hosting, which would go right to your bottom line profit. A couple hundred dollars a month isn't huge, but at this point in your business that's say $2,400 a year. From the looks of it, that's at least 2-3 months worth of revenue or almost 5 months worth of profit.
I think it's at least worth considering with where your project is at right now.
A factor that's usually forgotten is the developer's time. App Engine saves you A LOT of time compared to other solutions.
* Once you write your app the App Engine way, it scales automatically. You don't need to re-write periodically to accommodate growth, add replication, shard your data, ...etc.
* Data is replicated to several servers and data centers. You don't worry about losing data due to a hard drive crash.
* If an instance crashes, or a whole datacenter dies, your app keeps running.
Basically, measure how much time you send on non-code activities. App Engines saves you 90% of that.
"The App Engine way" involves writing a lot of things from scratch that normally you wouldn't have to, and spending a lot of time on kinds of optimization that you are unlikely ever to need. This offsets that perceived benefit in terms of developer time.
You could argue that it's good to spend this extra time because it makes your site more scalable. But if you ever reach the scale "the App Engine way" becomes relevant at, you are likely paying an order of magnitude more than you would if you bought dedicated servers. For all smaller cases, you are prematurely optimizing for scalability.
App Engine's reliability is easily overstated. App Engine has had a long history of issues with their services like datastore. When datastore does not function, there isn't any reasonable way to build a backup storage system because the platform is so restrictive. Frankly, backup services are cheap, and there are plenty of hosted database services if that's all you need. Ones which don't lock you in forever.
So, yeah, sites built on App Engine do go down for various reasons and still have to be monitored just like anything else. Except that you can't use any of the usual tools because the platform is completely unique and the underlying software is proprietary and completely secret to you as a platform user.
App Engine apps realistically only run on Google's servers. So even in the best case, you must have a lot of pure faith in Google. Because if you have reliability problems, problems with the platform's restrictions, problems with the terrible support or if the price skyrockets again, or if the platform gets wound down... you will be absolutely forced to rewrite your whole project to move it off of App Engine. You really won't have any idea that this is going to happen until man-years have already been invested.
It's great for standalone, spare-time projects that don't need 100% uptime or monitoring, though. I have low-traffic websites I haven't touched in months or years and they're still running fine. On anything else I would probably have shut them down by now due to some forced upgrade or other.
You pay a bit more developer effort up front (weeks, not years, for these projects) so you can walk away, but that's when I'm actually interested in working on the code. Migration is a non-issue since I don't plan to migrate. If I have to do any real work at all, I'll probably shut the site down anyway.
Yeah, I totally get that, but I guess I don't know what part of the Google AppEngine data store that works better or is a better fit than Amazon's DynamoDB, or running your own MongoDB or CouchDB or Hbase or Cassandra or whatever.
What does Google's datastore do that makes it worth sticking to and paying more for.
Excuse my naivety but isn't data store just a managed nosql database? You could switch over to Google Cloud (AWS but from Google) and still use Data Store or you could switch to any PaaS and use something like mongohq.
> Excuse my naivety but isn't data store just a managed nosql database?
Yes, so what? Its still something you need to cost out for an alternative.
> You could switch over to Google Cloud (AWS but from Google)
App Engine is part of Google Cloud (the AWS-equivalent umbrella of offerings). You probably mean "Google Compute Engine" instead of "Google Cloud" and "EC2" instead of "AWS".
> and still use Data Store
You could (Google Cloud Datastore seems to be very much the App Engine datastore outside of App Engine) but its then a separate cost you have to include in the comparison.
> or you could switch to any PaaS and use something like mongohq
You could, but its then a separate cost you have to include in the comparison.
Without knowing what your server load looks like, I would imagine you could save a couple hundred dollars a month in hosting, which would go right to your bottom line profit. A couple hundred dollars a month isn't huge, but at this point in your business that's say $2,400 a year. From the looks of it, that's at least 2-3 months worth of revenue or almost 5 months worth of profit.
I think it's at least worth considering with where your project is at right now.