I'm sure many of us have used PaaS cloud providers such as Engine Yard or Heroku as I certainly have - if you're using Rails that is. Now that they've been around for a few years and they can be evaluated better I can't help but wonder about the future of some of them and their long-term viability.
Starting off with Engine Yard, while what they provide is certainly useful and there are goodness knows how many man hours in that system, when it boils down to it all they're providing is an interface around Amazon EC2 (their web UI is really quite poor) and an integration framework. Although they provide private hosting as well, did this really need close to $40 million worth of funding? Seriously? Investors don't put that in out of the goodness of their hearts and one wonders how long they will wait to get it back.
When you hear of companies like this supposedly being sold to VMware, which didn't happen it, and at least one of the founders jumping ship, sorry, leaving, it has all the hallmarks of investors wanting to cash out. My opinion hardened further when I saw EY stopping you from using single instances for single servers, you have to use the medium instances now. No micro instances either. They're trying to seemingly expand into PHP hosting as well. I stopped using them when it became clear it was going to take them forever to get EU and Postgres support, which they now finally have. I had a rep from them 'reach out' to me literally every five minutes and I asked him over and over when this stuff would be ready.
On to Heroku, and the great thing about them is that it's very easy and fast to start running an application, and very cheap if it's small and useless. The downsides? It gets expensive very quickly and you are very limited with the applications you can run. Really? You're going to charge me all that money for me to squeeze my app into a read-only filesystem, extra for background jobs, no Varnish(?!), regimented software and no MySQL (not that I like it)? Then there's the add-ons............. Again, it's still only a platform built on EC2, so was it really worth the $212 million Salesforce.com paid? Their investors cashed out, and they only invested $13 million, I think, which I think is still way too much. Don't expect any price decreases any time soon, certainly not with SFDC.
It could just be cynical me, but I just think that these platforms built on top of IaaS environments like Amazon are very overvalued, expensive as a result for what they actually are and are focusing on the wrong things. What I'd be looking for is a cross between Heroku and EY with some really useful services such as proper monitoring, done on a budget without any financial silliness that acts like an elephant on its back. I'm still looking...........
Starting off with Engine Yard, while what they provide is certainly useful and there are goodness knows how many man hours in that system, when it boils down to it all they're providing is an interface around Amazon EC2 (their web UI is really quite poor) and an integration framework. Although they provide private hosting as well, did this really need close to $40 million worth of funding? Seriously? Investors don't put that in out of the goodness of their hearts and one wonders how long they will wait to get it back.
When you hear of companies like this supposedly being sold to VMware, which didn't happen it, and at least one of the founders jumping ship, sorry, leaving, it has all the hallmarks of investors wanting to cash out. My opinion hardened further when I saw EY stopping you from using single instances for single servers, you have to use the medium instances now. No micro instances either. They're trying to seemingly expand into PHP hosting as well. I stopped using them when it became clear it was going to take them forever to get EU and Postgres support, which they now finally have. I had a rep from them 'reach out' to me literally every five minutes and I asked him over and over when this stuff would be ready.
On to Heroku, and the great thing about them is that it's very easy and fast to start running an application, and very cheap if it's small and useless. The downsides? It gets expensive very quickly and you are very limited with the applications you can run. Really? You're going to charge me all that money for me to squeeze my app into a read-only filesystem, extra for background jobs, no Varnish(?!), regimented software and no MySQL (not that I like it)? Then there's the add-ons............. Again, it's still only a platform built on EC2, so was it really worth the $212 million Salesforce.com paid? Their investors cashed out, and they only invested $13 million, I think, which I think is still way too much. Don't expect any price decreases any time soon, certainly not with SFDC.
It could just be cynical me, but I just think that these platforms built on top of IaaS environments like Amazon are very overvalued, expensive as a result for what they actually are and are focusing on the wrong things. What I'd be looking for is a cross between Heroku and EY with some really useful services such as proper monitoring, done on a budget without any financial silliness that acts like an elephant on its back. I'm still looking...........