All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Best alternatives to Amazon EC2
Yes I googled "best alternatives to Amazon EC2" and a few variations. So I have seen a lot of those hits.
I'm a dedicated server guy. Don't know much about cloud services and their benefits. I played around with Amazon EC2 a few times but it's been a couple years. Thinking of moving some critical things like my WHMCS billing and Solus control panel to a cloud service.
Amazon seems a bit expensive once you add in bandwidth etc. and I'm not sure if it's worth it. Any alternatives people can recommend?
I just spent about an hour looking around and don't see many mature alternatives. A service like Digital Ocean or Vultr is not the right type of service for mission critical imho. So they are out.
Softlayer is really expensive and not sure it's worth it. Rackspace looks like it's worth a look. LeaseWeb also has something but don't know much about it and their European centered support kind of turns me off. Anything else I should be looking at?
Comments
Microsoft Azure. They do Linux.
Google Cloud.
But both are priced like EC2.
Alternatively, something like WiredTree or KnownHost.
what features of EC2 are important to you?
Also OVH. I still see them as a budget provider though. Not the sort of place for mission critical. That's for dedicated. Haven't heard much about their cloud service. For mission critical support is important so I want 24/7 phone or chat.
That it stays online basically. They seem to have the bandwidth to absorb DDoS attacks so that's a plus as well. Also might make use of the ability to scale up cpu/cores on demand.
With dedicated, sooner or later these guys all go offline for one reason or another and sometimes they are slow to respond. Amazon seems more solid in that respect
M$ Azure, Google cloud, hp cloud
Can I run a LAMP stack on Google Cloud? Last time I looked at them they didn't really do that. It was more about piecing these things together and paying for each individual thing. So more like Amazon AWS I guess.
Tbh, AWS has gone offline a few times in the past, and is not immune to outages. The same applies to all other cloud providers; there is no "magic" that makes the road construction crew not dig up their fiber optics/etc.
Have you thought of scaling out, and HA-ing everything instead of finding a single provider for a mission critical site? You can get around ~3/4 servers at various reputable providers around the US or Europe, and easily create a HA setup that will continue running if a few of the servers drop out.
keyword "Compute Engine".
btw how much your budget? recommendation depend on your budget
No way to HA WHMCS or Solus.
Well technically it may be possible to create a Galera Cluster for WHMCS but haven't tried. Solus tells me it's not possible to do anything like that with their control panel.
It's a big enought PiTA to update one WHMCS. Wouldn't want to have to do it 3 times.
I hear ya. I normally thought that too. However, I think cloud providers like Amazon are more well financed and have a lot more technical support. I am finding a lot of the dedicated server guys are not just bare bones servers but bare bones support. They don't have the kind of money that Amazon and Google has to throw into infrastructure and support.
I know Amazon had a few high profile outages a few years ago. Haven't heard of anything recently with any of them.
In this case, was going to suggest ScaleMP, but probably overkill for Solus/WHMCS.
So is that it? MS Azure and Google Cloud? Looked at Google Cloud and still doesn't look like I can do anything with a LAMP stack with them. They seem to be going for stuff that requires extreme scalability.
I don't want to do business with MS so won't consider them.
Er, what exactly are you looking at? If you want the equivalent of AWS EC2, then see https://cloud.google.com/compute/
I looked a google cloud and it looks like maybe overkill for what I want. Unless I am missing something. Google seems to be going for large enterprise. Maybe I'm wrong but that's the impression I get from looking at their site.
+1 for Azure if you need a solution that have about the same price than AWS. Take a look at OpenShift by RedHat, Google Cloud, Rackspace Cloud and IBM Cloud . They are all pricey and "enterprise grade" but have the best service level out there.
Ok thanks. That's the impression I am getting. I would rather do business with Amazon than Microsoft all other things being equal. I will take another look at Google Cloud too I guess.
Don't Google cloud and Azure both get a lot more outages than Amazon though?
Never got a outage using Azure but, yes. On the general case it's actually true.
Maybe you should consider OpenShift, they have quite a good price.
EDIT: and Softlayer, ofc.
Thanks for the suggestion. I will take a long look at them I think. If they check all the boxes then that might be the way to go.
Apparently Openshift runs on top of Amazon EC2. So the only benefit compared to Amazon as far as I can tell is that I get better pricing because they probably negotiate a much lower rate.
https://www.openshift.com/dedicated/
I think that if you're using AWS EC2, the OpenShift Online will be more suitable for your needs (https://www.openshift.com/pricing/index.html). They have a really good uptime when compared to "direct" EC2 and their infra will give you peace of mind about EBS data corruption and VM “retirement” (the actual lifetime for EC2 machines is of 200 days, less than a year).
@Sman I keep my critical stuff with stormondemand.com. Their network is clean, servers are fast, and although I don't ask for help, I am able to get a hold of someone within 15 minutes.
Is an AWS style service really what you're looking for or is it the perception of above average stability? I don't see you mentioning features too heavily, and your words imply to me that stability is really the key.
If that be the case, and you prefer single LAMP servers, I would propose that AWS is not going to provide you with any more stability than anything else. It's the features that allow you to build high availability applications that make it any more stable than say, RamNode or BuyVM. They still have servers go down, still null route heavy attacks, arrays still die. There's nothing fancy at the single hypervisor level.
Support is important, but when things go down I would rather a status page than someone over the phone telling me that they don't have a magic "fix" button because, quite frankly, they don't. The ability to hear a human voice doesn't fix that computers fail, or speed up what anyone has to do to fix them. You'll pay out the ass for guaranteed instant access to support too, and they still won't do anything because they're just support...not sysadmins or datacenter techs.
Just my two cents. I'm all about quality service at any cost, but no point in throwing away money if it doesn't come with a real, solid benefit.
To answer your question, Google and Azure. Google is actually very similar and a bit cheaper for the same performance, although plans change often.
In my opinion, the main advantage of AWS etc. is not in EC2 but the ability to use the other services: SQS, RDS (and Aurora), Route53 etc. If you really have no intention of using them then you'd be better with OVH's public cloud which is really, really good purely from a performance/price/reliability point of view.
I stared at those OVH plans for about 30minutes and couldn't figure them out. I honestly didn't know what I was looking at. They have storage plans and CPU plans and memory plans but I don't know why or what the differences are. I always found their packages and plans and terminology confusing.
Get 2 Ramnodes and be done with it.
They really add $20/mo to your bill for CentOS 7?
https://www.stormondemand.com/pricing/
aws gouges you on bandwidth all across their portfolio .
It's only worth it if your organisation requires extreme agility, like a unicorn startup.
You can run a failover type setup on Vultr and DO (i don't remember if both of them have unmetered private networking)
At the minimum, they both have snapshots - manual and automatic.
Kubernetes and other GIFEE projects are very interesting.
GIFEE = Google Infrastructure for Everyone Else
@sman What you do need to understand is
cloudreally just means a virtual machine with fancy controls and unlimited resources you can pay for that will never go down (or at least EXTREMELY unlikely). No cloud will restrict or care what you run assuming its legal and your paying.There is IaaS, Infrastructure as a Service (vm in the
cloud), then there is PaaS, Platform as a Service. PaaS is things like heroku, AWS beanstalk and wpengine.The scaling and HA is done as you have the tools to allow it in cloud, but your not forced to use it.
So at minimum treat AWS as a pay-as-you-go VPS provider or dedi provider, as it comes down to being the same thing, but with a ton of abstraction.