Why Cloud Sucks..

Cloud has obvious benefits for medium to small businesses but I'm a n00b in that environment. Hell the last contract I worked on as an IT operations admin had just short of a 1000 domain controllers and was brought down from over 10000 to 6000 servers. Half of which was VM's in their own data center with proper DR built in place. Good luck pushing all that into the cloud.
 
Sonic2k said:
Who the hell migrates their whole infrastructure into the cloud? Hence my ignorance at what he meant, but he is out of his depth on a lot of points. Saying that I don't own the software, well, I get presented with an instance, take for example EC2 in AmazonAWS, its a bare bones vanilla Linux box, I then have to install my stuff, and make it run. My stuff, as in my own applications, be it open source or ones I wrote and compiled in GCC. I have even gone as far as to compile from source, on one of these things. I can't remember what it was for, but it was for a daemon I needed to dump all incoming connections from the embedded devices in the field.
Click to expand...

Netflix migrated their whole infrastructure to the cloud. That's just one example of hundreds.
 
All your points are valid, I just want to state that yes, AWS had to reboot 10% of their fleet, but it was done in a rolling fashion, never the same Availability Zone at a time, therefore if you have architecture that is designed to be multi-availability zone (it's DR for those who don't know) which for a production service you should be, then you experienced no downtime during the 10% reboot. All availability zones would have to be down the EC2 SLA to be impacted. According to cloudharmony the EC2's uptime was 99.9984% this last year. That Good luck in achieving that in your own DC.

https://cloudharmony.com/status-1year-of-compute-group-by-regions

BTW the 10% reboot was to apply a critical patch in Xen to protect AWS customers. It was not actual downtime due to an outage.
 
psypher246 said:
All your points are valid, I just want to state that yes, AWS had to reboot 10% of their fleet, but it was done in a rolling fashion, never the same Availability Zone at a time, therefore if you have architecture that is designed to be multi-availability zone (it's DR for those who don't know) which for a production service you should be, then you experienced no downtime during the 10% reboot. All availability zones would have to be down the EC2 SLA to be impacted. According to cloudharmony the EC2's uptime was 99.9984% this last year. That Good luck in achieving that in your own DC.

https://cloudharmony.com/status-1year-of-compute-group-by-regions

BTW the 10% reboot was to apply a critical patch in Xen to protect AWS customers. It was not actual downtime due to an outage.
Click to expand...

Another example when Heartbleed hit AWS patched their ELB's (Elastic Load Balancers) before we could patch any of our inhouse stuff and we patched the same day it was released. So I would be very careful to say that "Cloud is less secure" its how you deploy to it...
 
Mweb's piece of **** vm offering isn't cloud though....
Its the same as when their red friends decided to lable their sub par virtual machines as cloud and also lost a ton of data, anyone remember that? The large ISPs are a joke on a race to the bottom with incompetent staff (disagree? try and get Afrihost to deal with an authentication problem into their DNS cluster, if they put all of the engineers I have ever spoken to together they barely have the brainpower to comprehend the subject never mind support it), nothing good can of this.
 
You cannot compare the local offerings from these greedy and inept ISPs to Amazon AWS.
 
Irritating headline/angle by OP.

The failures have zip zero nothing nada to do with "cloud".

At the end of the day, "cloud" is just a different route to your apps and data, hopefully to a DC that has top-notch systems, processes and practices managed by competent people who understand the risks and work smartly to mitigate them.

People and systems will fail. We know that for sure. Which is why anything less than three backups is reckless, with at least one off-site in case of physical systems loss (fire, theft, natural disaster, power failure, riot, etc).

Posting from phone so supa-brief.
 
Moosedrool said:
Cloud has obvious benefits for medium to small businesses but I'm a n00b in that environment. Hell the last contract I worked on as an IT operations admin had just short of a 1000 domain controllers and was brought down from over 10000 to 6000 servers. Half of which was VM's in their own data center with proper DR built in place. Good luck pushing all that into the cloud.
Click to expand...

I wouldn't say that some of these guys are small: http://aws.amazon.com/solutions/case-studies/all/
 
Ipwn 4 said:
vm offering isn't cloud though....
Click to expand...

Hosted vs 'true' cloud

http://www.netsuite.com/portal/resource/articles/true-cloud-solutions.shtml

Good resource. We also need to consider IaaS and PaaS, and vm instances which are true cloud.

The backup aspect, the additional security measurement, is the customer's responsibility to protect against events which aren’t covered in the SLA. However the Mweb Business incident is related to a human error by implementation which I would duly added to be in a most basic SLA as to be the provider’s responsibility. Backup or no backup, the environment exposed everyone utilising the service to the risk losing access to their data or losing their data. Then there is the pro-active support which is line with availability which didn't meet the standard.

The very same when comparing true cloud to simple SaaS. There is a very good reason why hybrid solutions are gaining momentum.
 
psypher246 said:
Netflix migrated their whole infrastructure to the cloud. That's just one example of hundreds.
Click to expand...

Indeed... I meant that sarcastically now that I re-read my post. I am very pro-cloud actually, once I got to see the benefits thereof.
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X