Reliable Cloud host ?

Jason Ackley jason at ackley.net
Tue Feb 28 02:50:25 UTC 2012


On Sun, Feb 26, 2012 at 4:56 PM, Randy Carpenter <rcarpen at network1.net> wrote:

> We have been using Rackspace Cloud Servers. We just realized that
> they have absolutely no redundancy or failover after experiencing a
> outage that lasted more than 6 hours yesterday. I am appalled that
> they would offer something called "cloud" without having any failover at all.

 Disclaimer: I work for Rackspace in a network architect capacity. We
have plenty of redundancy where it is needed.  We have all sorts of
solutions, for all sorts of intersections of problems, budgets and
customers. Sometimes finding the 'correct' solution is not as easy as
it could or should be. The menu is simply getting crowded :)

I don't know the specifics of your issue, but if you contact me
privately I can look into the specifics. You can also use my work
email address if you don't think I am legit (email me at this address
to get it). I do find that that impact is quite extreme, and certainly
an exception and something that there are many folks probably still
working on root causes and lessons learned. We take this stuff
seriously.


> 1. Full redundancy with instant failover to other hypervisor hosts upon hardware failure (I thought this was a given!)

As others have mentioned, you will not be able to find some of these
features for 1.5c/hr.   They quickly spiral out of control for
large-scale deployments. Every penny matters at 1.5c/hr .

I would ask that you look in your product portfolio and see if you
have anything at that price that you can answer a support phone call
for :) . This is not meant to be antagonistic, just to have a clear
mindset understanding of the $$ we are talking about and how careful
you have to be.

What these cloud price points allow you to do tho is to turn it from
one type of a problem to another type of problem that you can have
more control over.  As others have mentioned, spreading out with many
different providers is one example.

They (cloud, VMs, VPS, whatever you want to call them) are cheap,
disposable computing resources - don't treat them as anything else! As
with anything, you get what you pay for, and I am sure we have all had
'that customer' that claims $1,000,000 in losses for every hour of
impact, and they have a single whitebox server deployed.


> 2. Actual support (with a phone number I can call)

This is where the providers will typically start to differentiate
themselves from each other. As a company, we pride ourselves on
support. Full support has a price.  I don't want to turn this into a
sale-ish email tho.


> 3. reasonable pricing (No, $800/month is not reasonable when I need a tiny 256MB RAM Server with <1GB/mo of data transfers)

1.5c / hr is what our basic linux image starts at IIRC. Again, I am
not in sales, so I don't really keep track of how that compares to
some of the other folks out there, I would guess it is about the going
rate.

I have used Linode.com as well as EC2 as well and they both have some
great feature sets and offers. Both also have areas that could use
improvements.

I do agree that there is general misconceptions of what 'cloud' means.
That is simply a byproduct of the amount of folks involved in such
trends, and yes, the marketing folks getting involved as well. This is
unavoidable in the world today.

If you have any other questions or concerns that I can help with,
please let me know...


cheers,
--
jason




More information about the NANOG mailing list