Why We Moved Off The Cloud • Hacker News Board

Re: Why We Moved Off The Cloud

#141

Post by chrissnell » Fri, Oct 28, 2011, 7:21 PM UTC

powertower wrote: post #97

jwegan wrote: post #63

Rackspace's primary business is dedicated hosting. In fact they let you have both dedicated hosts and cloud hosts and allow dedicated and cloud hosts to talk to each other. I'm curious why they didn't just switch to Rackspace's dedicated hosting. It would have given them the performance they needed while retaining the flexibility of being able to quickly spin up cloud machines in the same datacenter as the dedicated…

> In fact they let you have both dedicated hosts and cloud hosts and allow dedicated and cloud hosts to talk to each other. That's not amazing at all. You do realise that a "cloud" host and a dedicated box are two exact same hardware boxes siting next to each other in a rack? One's just virtualized 10x with Xen, VMWare, KVN, etc.

You're wrong about that. The big hosting providers do not put public cloud hardware next to dedicated hardware. Public cloud hardware goes in big, homogenous racks in a specific area of the datacenter. Dedicated hardware is very mix-and-match; you have all sorts of servers and network devices racked together. Connecting the two environments is no small feat, given the VLAN limitations of most current switches.

Re: Why We Moved Off The Cloud

#142

Post by noodle » Fri, Oct 28, 2011, 7:59 PM UTC

crag wrote: post #64

noodle wrote: post #43

Earlier quoted context omitted.

those are examples of people making (or being sold on) poor choices and/or not having all information necessary to make an informed choice. the cloud is not for everyone. physical servers aren't for everyone, either. stories like these don't automatically imply that using the cloud is a bad idea for everyone.

"informed choice" No, no. It's about price. Most small business buying subscribing to these services aren't tech. In the phone market (as an example), carriers are selling hosted "solutions" for less then $75 a month. Comcast is too. I like Comcast. But you can't run an office with 25 phones and pc's and other devices on Comcast. At least not in Florida. The other problem, is most IT firms down here are pushing they…

if it was purely about price, then these people would've purchased these options regardless of the shortcomings if they had known about them beforehand. and if this is the case, they went in with eyes wide open and the cloud is right for them.

Re: Why We Moved Off The Cloud

#143

Post by crag » Fri, Oct 28, 2011, 10:07 PM UTC

noodle wrote: post #142

crag wrote: post #64

Earlier quoted context omitted.

"informed choice" No, no. It's about price. Most small business buying subscribing to these services aren't tech. In the phone market (as an example), carriers are selling hosted "solutions" for less then $75 a month. Comcast is too. I like Comcast. But you can't run an office with 25 phones and pc's and other devices on Comcast. At least not in Florida. The other problem, is most IT firms down here are pushing they…

if it was purely about price, then these people would've purchased these options regardless of the shortcomings if they had known about them beforehand. and if this is the case, they went in with eyes wide open and the cloud is right for them.

I doubt that. Trust when I tell you this, the carriers (and I'm including Comcast) sales people do not tell the customer the downsides and limitations.

If you don't believe me. Try it. Call the business sales units of the carreirs.

Now I believe in buyer beware. But that's the problem... most small business are suffering cash flow problems. When something cheaper comes along there is no "buyer beware". They think about lowering their monthly bills. An of course, in the end they get bit in the ass.

It's just mind-blogging to me how many business owners are so ignorant about the tech that runs their business.

Re: Why We Moved Off The Cloud

#144

Post by seunosewa » Sat, Oct 29, 2011, 12:07 AM UTC

tzury wrote: post #32

Facebook's CTO thinks the opposite or perhaps even alike. IMHO, a startup should not start with dedicated, but perhaps when you get to a certain size, perhaps, dedicated hardware, team and bandwidth is the way to go. Just like you will not have a dedicated chef to cook your meal at start, but rather outsource meals, etc. -- http://www.bbc.co.uk/news/business-12406171 What's the biggest technology mistake you've ever…

What about dedicated servers?

Re: Why We Moved Off The Cloud

#145

Post by rachelbythebay » Sat, Oct 29, 2011, 1:14 AM UTC

asharp wrote: post #134

rachelbythebay wrote: post #91

Earlier quoted context omitted.

Beware the "same config everywhere" approach. It works up to a point, then it turns into a disaster. All you need is one totally broken change like "chmod -x /usr" to really make life interesting. You start bleeding machines and pretty soon you have nowhere left to host your tasks. It's interesting, right? At first, you can handle a couple of totally mixed-up machines. Then it stops scaling and you have to start doin…

You just need to do rolling deployments to make simple things like that go away. Roll out a deployment to N machines (like say 10), run self checks (you have those, right?), if everything passes give them standard load. Over the next K period of time, periodically check up on them. After that, roll over to N*2 or N^2 nodes, continue until you have rolled out to your entire cluster.

Some overly-confident human went "oh, this can't be bad" and force-approved it. This really happened. Multiple times. And they have rolling deployments.

Their response? More administrative measures.

Re: Why We Moved Off The Cloud

#146

Post by necro » Sat, Oct 29, 2011, 1:34 AM UTC

Huppie wrote: post #126

necro wrote: post #114

Earlier quoted context omitted.

I have to agree. I put our stuff in a colo 2 years ago and never looked back. Pretty much all servers come with some kind of remote console interface IPMI, and that's not terminal redirection, thats actually a totally self contained microprocessor and ether port that you can run on a separate subnet and control your server even if it's off. I updated the bios, reinstalled OS's, all via IPMI which is part of the mothe…

I have NEVER had to drive/fly to fix anything. Never even had to use remote hands for anything. Sure some drives died, but standby drives are in place. Consider yourself lucky. We thought the same thing, but when a RAID controller died on us recently we really didn't know what hit us. It didn't just stop working, it started by hanging the server every now and then, then after a day slowly corrupting drives, then afte…

Im a bit conservative when it comes to hardware like raid controllers. My choice was 3ware. They are by no means the fastest, in fact the performance sucks compared to others. I went to a company that builds storage systems, but will build any kind you want, not locked into any controller. I trusted them when they recommended that by their experience is returned/fails the least. Of course everything fails, so it's just a matter of time. We have tripple redundant storage for file backup... active, 5 minute backup that is ready to be swapped in at one click, and long term. If something goes wrong with the active set or slows down, we just flip a switch and all our app servers use the new system that at most is 5 minutes behind. Old system gets shot in the head, and can be diagnosed off line. Shoot first ask questions later.

Re: Why We Moved Off The Cloud

#147

Post by pbreit » Sat, Oct 29, 2011, 6:16 AM UTC

powertower wrote: post #97

jwegan wrote: post #63

Rackspace's primary business is dedicated hosting. In fact they let you have both dedicated hosts and cloud hosts and allow dedicated and cloud hosts to talk to each other. I'm curious why they didn't just switch to Rackspace's dedicated hosting. It would have given them the performance they needed while retaining the flexibility of being able to quickly spin up cloud machines in the same datacenter as the dedicated…

> In fact they let you have both dedicated hosts and cloud hosts and allow dedicated and cloud hosts to talk to each other. That's not amazing at all. You do realise that a "cloud" host and a dedicated box are two exact same hardware boxes siting next to each other in a rack? One's just virtualized 10x with Xen, VMWare, KVN, etc.

Who said it was amazing? And I highly doubt that virtual and dedicated are the same boxes sitting next to each other (or even near each other).

Re: Why We Moved Off The Cloud

#148

Post by pbreit » Sat, Oct 29, 2011, 6:20 AM UTC

powertower wrote: post #11

> One last thing about getting dedicated hardware. It's cheaper a lot cheaper. We have machines that give us 2-4x performance that cost less than half as much as their cloud equivalents and we're not even co-locating. I've been saying this for ages, and every time people would fall over backwards trying to defend/prove their cloud mistake... "The cloud is cheaper, faster, and infinitely scalable." Except none of thos…

I don't think I've ever heard anyone claim that the cloud was generally cheaper and faster. Only that it is cheaper at small scale and faster to set up.

Re: Why We Moved Off The Cloud

#149

Post by proofpeer » Sun, Oct 30, 2011, 12:30 PM UTC

cr4zy wrote: post #140

proofpeer wrote: post #15

I think you have to distinguish between where you rent whole server capabilities, and more managed models like Google App Engine. It seems to me that GAE should deliver more consistent performance.

GAE definitely does not deliver more consistent performance. I've done a ton of performance tweaking on GAE as of late and my bottlenecks are now reduced to random points in the code between RPC's where my Python thread is obviously locked out of the CPU it was running on. I should say that the variation this leads to is at max around two seconds. I believe this is due to App Engine doing some dynamic grouping of slo…

That's interesting. Do you think you get better performance by using Java instead of Python? I heard bad things about Python and the GIL.

Re: Why We Moved Off The Cloud

#150

Post by jwegan » Sun, Oct 30, 2011, 6:05 PM UTC

powertower wrote: post #113

jwegan wrote: post #106

Earlier quoted context omitted.

Of course its not amazing, but it is a feature they have that would have made the transition to dedicated easier.

Being networked is a feature?

Being in the same datacenter is a feature since it means there is extremely low latency between your dedicated and cloud machines. Sub-millisecond vs. 10s of milliseconds round trip times can make a huge difference.