Can the Cloud Be More Secure Than Your Own Servers? (Video)
Sarah Lahav, CEO of Sysaid, believes "the cloud" can be more secure than keeping your software and data behind your firewall and administering it yourself, especially for small and medium-sized firms. Why? Because Amazon, Rackspace, and other major cloud and SaaS providers probably have lots more security experts and other IT people at their command than you do.
We've talked to Sarah before, and probably will again. She has strong opinions based on her experience in IT, and is happy to share those opinions. So take it away, Sarah...
We've talked to Sarah before, and probably will again. She has strong opinions based on her experience in IT, and is happy to share those opinions. So take it away, Sarah...
Next question.
"...probably have lots more security experts and other IT people at their command than you do" well, i'm convinced ... here's all my data!
Can the cloud be more secure than your own servers? Yes.
Can the cloud be less secure than your own servers? Yes.
Guess what it costs me to have a connection so stable that it never goes down?
As it turns out, it is far more (measured over 5 years, the length of our ISP contracts) than proper redundancy in my equipment costs.
Amazon, Rackspace, et-al don't give a shit about your data.
They care about the data your data generates. That is backed-up, carefully guarded and controlled. Your data on the other hand, it stored on the B and C grade disks, tapes and run on any old CPU in the farm that is past it's prime.
Centralized data is great, for hackers. One target, lots of data, lots of reward. Targeting that one user, with the firewall? Not so much.
While a cloud server has more security resources, they also have more professional hackers targeting them, since a single exploit has a good chance of bagging all the cloud provider's customer data. Think attacks like the Sony breach were bad? Just wait until you can get Sony, Microsoft, Facebook and the state of Ohio all at once because they happen to be hosted by the same cloud provider.
OTOH, perhaps that might just be the best place to be when a zero day drops. A cyber criminal won't likely bother with a small business and just go straight for the 23 terabytes of customer data on the next rack over...
HA! I just wasted some of your bandwidth with a frivolous sig!
Somebody flashes a badge, and they just hand your shit over, no questions asked... if they know what's good for them.
“He’s not deformed, he’s just drunk!”
Where have the past 2 years major data breaches occurred: Off-Cloud.
But what about adjusting for Cloud vs Off-Cloud %-usage: Still no contest.
Science & open-source build trust from peer review. Learn systems you can trust.
Most drivers consider themselves to be above average. Why would that not extend to server operators?
Is it just my observation, or are there way too many stupid people in the world?
Because Amazon, Rackspace, and other major cloud and SaaS providers probably have lots more security experts and other IT people at their command than you do.
But those experts aren't regularly upgrading software I run on their cloud systems to fix security holes, nor monitoring my sites for exploits. So their expertise buys me little--other than the underlying infrastructure hopefully will be sound. That's all. That's not lot. The majority of security bugs/holes I've had experience seeing exploited were holes in application packages (think WordPress). Unless you mean hosting your resources on a specific application hosting provider who handles all upgrades (i.e. a hosted WordPress provider in this example, who guarantees up-to-date bug fixes on WordPress and some set of commonly used plugins).
If data is on my personal server and the US government wants to see it, they need a warrant.
If it's on a cloud server, they don't.
To a Lisp hacker, XML is S-expressions in drag.
Why is this taking megabytes of bandwidth to convey a message that could take kilobytes? Is there something visual about this concept that can't be communicated in writing? Stop the dumbing down of of /.
This is like saying that Budweiser has better beer than a local brewery because they have bigger vats and more distributors.
I think the trick to security is not in how many experts you have, but in how willing you are to cut corners to increase profits.
You are welcome on my lawn.
YOU CANNOT BE SERIOUS!!!!
She is the CEO of a cloud based company. What the fuck do you expect her to say?
The real question is not...is the cloud secure? The question is...who is more likely to be a target of hackers?
Can cloud services be made secure? Of course it can. But it doesn't necessary mean that it is. It all depends on policies and procedures which you, as an end user, have absolutely no say in. And what happens if there is a data breach? You get a year of free credit monitoring. Thanks for playing. There is no implicit guarantee, or liability, on their part.
If you are a hacker who will you target? Me - with maybe a few credit card details or Amazon with millions or credit card details. The answer is obvious.
When it comes to the cloud I am reminded of the Tony Montano (Scarface) quote: "Who do I trust? ME!".
Depends, do you have a dedicated security team?
The security grunts are paid in Alpo, and the supervisors are paid in Meow Mix. I also pay their medical.
Start with the fact that cloud services are big, ripe, juicy targets for anyone and everyone. Continue that there's probably never a time when their service isn't under some kind of attack in one way or another. Add in the fact that my server contains nothing of any real value to anyone but me. And extrapolate that to a very low likelihood that anyone would bother to take the time to attack my server. Consider also the fact that the cloud provider has to succeed 100% of the time to make my data secure while the hackers can fail almost forever and only have to succeed once.
I'm going to go with the fact that my data is more secure in my server at home than it would be in the cloud.
Of course, small businesses without a dedicated security teams are legitimate targets. But whether they store their data in the cloud or in company servers, their business internet connection is vulnerable to attack and provides a much easier road into the cloud storage than trying to directly attack the cloud servers. So realistically, the businesses accessing the cloud servers in bulk are a significant vector for attacking a cloud service. As a result, it doesn't matter where the business stores its data, it is no more or less vulnerable to attack in either location.
When it comes to large corporations, they are bigger targets but they have the budget to hire security experts just like the cloud provider has. So while they too are probably under constant attack 24/7/365, they are not necessarily any more or less vulnerable than the cloud provider.
So on balance, I'm going to go with no, the cloud does not necessarily make your data any more (or less for that matter) secure than not using it.
There's often a lot of focus on actual/active security, and a lot less on determining the need for that security. Think of security like a power-to-weight ratio for performance.
The goal isn't to have great security. The goal is to have no successful attacks. "no successful attacks" is approachable from two primary vectors: "successful" and "attacks". Security focuses on the successful vector, by resisting.
Certainly, when it comes to contracting a provider, or rolling my own, a big provider might be better than I am. Of course, I can hire a consultant and get the best of both, and a big bill to match.
Obfiscation is not security. But it is a reduction in the actual number of attacks -- so long as it's working, of course.
I've been with small providers, I've been with large providers, I've been with Rackspace, and I've rolled my own.
The truth is that all four scenarios have had plenty of attempted attacks. But dive a little deeper, and something way more interesting appears.
When I rolled my own, I got loads of random attacks, mostly from China. Nothing persisted for very long. Nothing was particularly focused. And nothing was complicated. Almost all were easily dodged with standard surface-area-of-attack controls, like closing unused ports and not having general server bloat.
When I was with Rackspace, I had loads of help from their excellent support teams, and on occasion, wow did I ever need it! Persistant attacks, lasting for days, targeted attacks, ddos attacks with large systems on the other end. At one point we had over a dozen rackspace support personnel just fighting to kill stuff fast enough to keep performance up long enough to identify and resolve the issue without needing to take the server entirely offline.
I was very happy with Rackspace, and was with them for a decade. Now I'm rolling my own again, things are just much more stable that way.
So what's your preference? Being in a military compound, protected by a thousand soldiers in the middle of a war-zone; or being completely unprotected, on a mountain side, in upstate montana?
I'm choosing big-sky country, personally.
Also, I believe that Rackspace is partnered with a very familiar government spy agency quite directly -- since they both moved campuses at the same time the other year, and I was greeted quite aggressively, as you would imagine, when I visited Rackspace for a tour, and accidentally pulled up to the unmarked neighbour. Probably appropriately so, given that it was on a september 10th.
Our company contracted with an external supplier to manage an application for us that we had been managing in house. We got the usual assurances about their data centre, nailed down the SLA, and did a PIA. All good. As we were working with them to get our data moved over one of our sysadmins came upon a SQL Server admin id/password, unencrypted, in one of their .ini files. It was pretty generic (the name of the application with a few numbers instead of letters). That looked suspicious to us, so we contacted another one of the same vendor's hosted customers and said,"I'll bet we can guess your SQL Server admin password in one try." Turned out they were using the same admin credentials for all their hosted customers databases. Which they kept unencrypted in an .ini file.
So yeah,maybe their data centre was secure, but their application level management was amateur hour. And it was a bit of a fluke we discovered that.
Needless to say, we never did move the application into the cloud. They promised to fix the problem when we brought it to their attention, but we didn't trust them after that. And even though they arguably violated the terms of our SLA, they were such small potatoes that there was no point getting the lawyers involved.
Problem being that if you do not take care of your security, you are also likely to not take care of your own security in your cloud instances.
For example, not too long ago some company got bit by leaving something wide open in how they set up their EC2 instances.
You cannot sprinkle on security as an afterthought either way, security is a factor that must be kept in mind as you do the design.
XML is like violence. If it doesn't solve the problem, use more.
Could you at least get some decent audio if you are going to do these? I listened to the first 10 seconds and could not stand the sound. Also you have a smart audience, we would much prefer to just read the story. We all thought the book was better.
Bunk. BS.
1) She has a vested interest in presenting that her systems are secure.
2) She offers a weak link in the data chain. Every time any link is added the system gets LESS secure. Adding a weak link further weakens the system.
Only non-secure data gets stored on the cloud. Remember, it's like a postcard.
I'll provide my own security.