Why we moved away from AWS • Hacker News Board

Re: Why we moved away from AWS

#21

Post by zerop » Mon, Sep 23, 2013, 2:37 PM UTC

I am planning to move from AWS to Linode mainly because of performance. My app is CPU intensive. I think for such apps you need to take high end EC2 instance.. I tried with small and medium instances but found them quite slow.

With linode 8 core small instances, I could handle 2-3 times the traffic. However from management perspective AWS rules.

Re: Why we moved away from AWS

#22

Post by InclinedPlane » Mon, Sep 23, 2013, 2:39 PM UTC

eterm wrote: post #4

EC2 was designed for elastic computing. On demand high computation (low memory) that are elastic. With that in mind, pure EC2 is a terrible choice for general web application hosting. If using the complete AWS set (S3, simpleDB, etc) then it makes more sense as stuff like db hosting can be pushed out to their services designed for it, but if you're gonna fire up a windows box, stick SQL server on there and use it as…

I know this is a tangent, but I think it's a worthwhile one to mention that backups and redundancy are not the same thing. There have been a few high profile ventures (including businesses) that had to shut down because they lost all of their redundant data in some way. Redundancy doesn't save you from malicious people who've gained access to your systems. It doesn't save you from errors (oops, dropped the wrong DB, thankfully it's .... replicated virtually instantly across all RAID volumes and clustered DB instances). It doesn't save you from the one building where all your data is in burning down or getting flooded. It doesn't save you from software bugs (either yours, in firmware, in the kernel, in the DB, etc.) from corrupting data.

Re: Why we moved away from AWS

#23

Post by solvemenow » Mon, Sep 23, 2013, 2:40 PM UTC

>there are also downsides when moving it to your server, more system administration, you have to build your own firewall, take care of security & backup, et

Startup idea right there. But then if I thought of it so quickly, somebody probably already does this.

Re: Why we moved away from AWS

#24

Post by LanceH » Mon, Sep 23, 2013, 2:41 PM UTC

There is a dead spot between using EC2 on demand and paying for the 3 year reserved instance, both of which I've found to be practical.

At both ends of that spectrum, however, I've found the pricing to be fairly reasonable. It just might not work for a startup.

Re: Why we moved away from AWS

#25

Post by himakara » Mon, Sep 23, 2013, 2:43 PM UTC

Nice post. It is important to note that these tend to be cyclical. As start-ups go through various stages of their life cycle, PaaS/ IaaS providers update their offerings and technologies mature/ invented, the appeal may shift between these options. I think it makes it even more important to build your technology stack in a way that is:

1) easy to deploy, migrate and update (using standard deployment technologies) and 2) least dependent on a specific vendor (GAE ;)

Re: Why we moved away from AWS

#26

Post by hashtree » Mon, Sep 23, 2013, 2:43 PM UTC

Compared to custom colocated clouds, you scale, code, and build your stack completely differently. I could not do half of what I do under any PaaS/SaaS.

I avoid disk at all costs (nearly unattainable amounts of RAM on PaaS/SaaS), if disks are hit they must be SSDs, treat everything immutably, concurrent/distributed computing, assume hardware is plentiful (192+GB ECC, 24+ of new xeon cores, etc). I scale completely differently than most. They really get you on RAM, I can build whole servers for what it might cost for a month of PaaS/SaaS.

Re: Why we moved away from AWS

#27

Post by cbg0 » Mon, Sep 23, 2013, 2:50 PM UTC

hashtree wrote: post #17

eminh wrote: post #3

I currently spend ~$2000 on Softlayer for six servers and use about 30TB of bandwidth. On AWS I would have paid more just for that bandwidth.

And you can pay much less than half that via custom server builds and colocation. It is just a matter of how far down the chain you want to go, given your expertise and sensitivity to hardware costs.

He doesn't actually have to keep replacement parts in the datacenter or have staff close by to the datacenter to go and perform replacements or new installs, or worse - pay >100$/hr. for remote hands with colo.

Over time it's certainly more expensive to rent, but you get to cancel and move on to better hardware when it comes out, without having to worry about re-purposing or selling old servers.

Re: Why we moved away from AWS

#28

Post by thomasmeeks » Mon, Sep 23, 2013, 2:50 PM UTC

thejosh wrote: post #15

leokun wrote: post #13

If you move to Rackspace, stay away from DWS, the dallas datacenter. It's over-booked, the network has constant issues, vm's on the same host machine as you are able to cause your vm network issues, the list of problems never stops. We recently switched to Azure from Rackspace, but we're still evaluating if it will work for us long term. Azure's issues are that you have to request number of core increases, and you ca…

How many server(s) did you have with RS? We only have 2 mid-sized virtual servers in DFW and things have been working flawlessly for us..

Really the issue is that virtual hosts can affect each other quite a bit on Rackspace compared to, say, AWS. If your server behaves poorly, Rackspace can and will shut it off. One of our non-critical servers ran out of memory, thrashed swap, and was shut down in pretty short order by Rackspace. Which is good, sort of, I don't want to hurt other customers. Still, getting it turned back on was not a very fast process.

So it is kind of a roll of the dice. Are the other customers on your hardware well behaved? Will they stay that way?

It is a trade-off, you get way better performance if the other virtual hosts on the box are quiet. But if you plan your capacity around those quiet periods you can be in for quite a shock once the hardware gets busy. I've run critical servers on hosts like this and it can be a headache.

Re: Why we moved away from AWS

#29

Post by leokun » Mon, Sep 23, 2013, 2:54 PM UTC

thejosh wrote: post #15

leokun wrote: post #13

If you move to Rackspace, stay away from DWS, the dallas datacenter. It's over-booked, the network has constant issues, vm's on the same host machine as you are able to cause your vm network issues, the list of problems never stops. We recently switched to Azure from Rackspace, but we're still evaluating if it will work for us long term. Azure's issues are that you have to request number of core increases, and you ca…

How many server(s) did you have with RS? We only have 2 mid-sized virtual servers in DFW and things have been working flawlessly for us..

It's because you only have 2 servers, which reduces the likelihood that you're sharing a host machine with a misbehaving vm. You're probably also not using a load balancer or taxing the network much yourself.

Re: Why we moved away from AWS

#30

Post by toomuchtodo » Mon, Sep 23, 2013, 2:56 PM UTC

InclinedPlane wrote: post #22

eterm wrote: post #4

EC2 was designed for elastic computing. On demand high computation (low memory) that are elastic. With that in mind, pure EC2 is a terrible choice for general web application hosting. If using the complete AWS set (S3, simpleDB, etc) then it makes more sense as stuff like db hosting can be pushed out to their services designed for it, but if you're gonna fire up a windows box, stick SQL server on there and use it as…

I know this is a tangent, but I think it's a worthwhile one to mention that backups and redundancy are not the same thing. There have been a few high profile ventures (including businesses) that had to shut down because they lost all of their redundant data in some way. Redundancy doesn't save you from malicious people who've gained access to your systems. It doesn't save you from errors (oops, dropped the wrong DB,…

This is why you make backups from your physical hardware to onsite storage, but also replicate those backups to Amazon S3 and inhibit the delete functionality so you need MFA in order to complete the delete.