Up to this point, Geni Whitehouse has introduced us to the cloud, its types, and different deployment models of Infrastructure as a Service. Then, she shared some valuable tips on how to choose the right cloud service provider in the previous video.
Adding to that, in the final video of the series “Journey To The Cloud,” Geni interviews Joanie Mann about her expert views on the critical attributes that a cloud service provider needs to have in order to provide optimum performance, data security, and service availability to the customers.
Joanie Mann has spent several years in domains like ISV cloud enablement and SMB application service delivery. She has been focusing on cloud hosting and online application service models for the past 20 years.
Let’s watch Joanie Mann’s interview by Geni Whitehouse –
Geni – In previous videos, we talked about what the cloud is, how to choose a service provider, and we touched slightly on data center security and disaster recovery. In this section, we want to talk a little bit more detailed about all of these three topics. And also, I have the honor of being joined by an expert who can really help us dig deeper into some of these topics.
This is Joanie Mann, who is joining us.
Joanie – Hi! My name is Joanie Mann. As Geni just said and I have been involved in application hosting for a lot longer than care to admit, frankly.
Gen – And I know Joanie for her bunny slippers. What’s that all about, Joanie?
Joanie – Bookkeeping in Bunny Slippers is all about being able to work anywhere, anytime, and nobody can see your feet on a conference call.
Geni – Bunny Slippers are cool. So, Joanie, let’s just jump right in here. So, when we are talking about data center, why should I care as a client of a solution provider? What do I need to know about that?
Joanie – O my goodness, there is lot of different aspects about the data center, and the data center represents, in some cases, the actual key to the capabilities of the service provider. You know we got different situations that we need to be really considering about, you know location, for example, being one of the big aspects of the data center.
What is the geography of the data center? Is it in a seismically active location where it is prone to the earthquake? Is it near water where it is prone to flooding? Is it in an area where it is prone to fires? So, the physical location, the proximity of that data center to risk areas, is really quite an important consideration. Another thing is we need to be concerned about how far the data center is away from us?
Geni – What does that do to me?
Joanie – Well, the longer the information has to travel the wire, the longer it’s going to take. You will get better performance if the data center facility is closer to you.
Geni – So, if I know that my service provider has multiple data centers, does that solve my problem?
Joanie – Simply having multiple data centers doesn’t necessarily solve the problem. That’s where we start talking about how the provider uses the data center. For example, are they replicating data only, using it for backup, or are they actually replicating entire systems so they can deliver full services from those different areas?
Geni – So I thought I just needed to worry about my data, but what I also need to worry about is everything else – all the applications that the data is running in and all of that stuff. I need to make sure that all of it is gonna be available when I need it.
So now, let’s move on from location, which is important. So, you are saying that the California earthquake zone is not my number one place to put the data center.
Joanie – Yeah, maybe not the best place to put it. (laughs)
Geni – So how do I dig into that when I am looking at a service provider? How do I make sure they will give me all that info?
Joanie – Generally, they will. You want to know where their physical locations are, where their data centers are located, and I mean you can do a little bit of research yourself. The kind to understand what risk those locations might potentially present?
You know but most service providers if they are using high-tier data centers, those data centers themselves are going to already have a lot of the properties the service provider needs in terms of the facility, providing redundant power, being able to provide redundant connectivity, many different areas of redundancy even within the data center facility.
So, the service provider is going to be able to take advantage of whatever it is the data center provides. So, a top tier data center is obviously the most desirable.
Geni – So Joanie, what should I look for in terms of performance when I move to a hosting provider? What should I expect, and what determines what kind of speed I am able to experience?
Joanie – Well, there is a lot of different factors that impact the speed, the performance of your cloud service. Whatever type of cloud service that is. One of the biggest initial impacts is your own internet connection.
Geni – I can’t blame everything on them? (laughs)
Joanie – That’s right, and that dial-up connection is not gonna do any good anymore.
So, besides having a great internet connection, having broadband high-speed Internet, one of the big keys is to make sure that the upstream (or the upload) and the download speed are roughly the same.
We call that being synchronous, and the big reason for that is because with our particular type of technology, the type of technology that the hosting providers are delivering, we expect a character (for example typed on the keyboard) to make a round trip to that server, show up (register) on the server, and then come back and show up on the screen immediately.
And if we don’t see that happen, that’s latency, and that type-ahead kind of situation, more often than not, that’s simply a function of your internet connection, maybe being able to download things more quickly than it can upload them. So, there is just kind of a disconnect in trying to synchronize that data, going back and forth.
Beyond your own internet connection, now we are talking about the things that your service provider control. And when it comes to the service provider, the big things that they have to concern themselves with are, for example, modern technology.
Computing technology has evolved, and so the service provider has to make sure that they are using modern equipment, modern operating systems, and they have to have a certain amount of agility to be able to upgrade those and refresh that equipment into maybe on a two-year cycle. Try to afford that in-house.
With the service provider, that’s one of the key aspects that they have to be using modern computing technology, and even more than that, as that technology evolves, they need to able to find ways to integrate that into their solution.
For example, instead of using spindle hard disks which used to be the way, you know, all the storage used to be on hard drives. Well, now we are talking about SSDs (solid-state drives) and flash drives, and you know, with no moving parts, those drives can operate so much more quickly, which increases application load time, it increases the data access times, and all of these different things that together represent the performance on the system.
Geni – There is a lot to consider when we are looking at that. It’s great that we don’t have to worry about it. We have to know that is the provider has taken all that into account if we have the right one. That’s good news.
OK Joanie, so you slipped a word in there I want you to draw into a little bit – Redundancy. What does that mean? That just not me repeating myself?
Joanie – No, well, redundancy is just that. It’s many replications of the same thing. So, when we look at the data center, which is obviously our topic of conversation, we are looking at all of the different aspects of redundancy even within that data center.
Now the service provider is going to have redundant equipment, redundant networking, and all those types of things. But when it comes to the facility itself, the building and the electricity – If they lose electricity, is there another source of electricity? If they lose whatever sources of electricity they have, do they have a generator? Is there fuel for the generator? Can they get more fuel for the generator? Internet connectivity- Do they have more than one provider? Because that will be very depressing for a cloud provider to lose the Internet.
Geni – Depressing for everybody. (laughs)
Joanie – You know, so we are talking about myriad levels of redundancy from the lowest level – two hard drives and a machine and something like that, all the way up to – not two copies of the building. Well, maybe we are talking about two copies of the building. This is why many service providers are going to have multiple locations because they are looking at redundancy not only within that particular facility but redundancy of their services across perhaps multiple facilities.
Geni – So, we have redundant hard drives, so I have hard drive at one place, and at second place, I have redundant machines, and now you are saying that we need to worry about redundant data centers.
Joanie – Absolutely!
Geni – Does everybody have that?
Joanie – No! Absolutely not.
Geni – I have multiple locations. Am I automatically redundant in terms of facilities?
Joanie – Well, you know, let’s maybe look at that as it compares to your own office. For example, if you have got multiple backups of your data that’s useful. But that doesn’t recover your system in its entirety. So, when you are looking at service providers, the same thing is true. There might be data replicated to another data center facility, but that’s not necessarily going to bring a customer back into production if one of those facilities goes away.
Geni – So, I need to really understand what they are saying when they tell me certain pieces of data because it doesn’t always mean the same thing.
OK, so there is this big topic. This thing that keeps us all up at night. What about security. What do I need to know and understand, and what am I looking for when I talk to my data provider, or when I am trying to choose one?
Joanie – Just trust me (laughs)
With security, it’s a big question. It’s a lot like redundancy; there is a lot of different pieces to it. We talk about the physical security in the data center – biometric controls, secure locks, all sorts of different things that are going to prevent unauthorized access into the facility, into the computers, etc. You know you do not want somebody walking into the room and kicking your server over or something like that, spilling their coffee, hitting it over with a vacuum, whatever.
But, in addition to the physical layer of security, now you have got the network layer, so we have got to transport the information that’s moving perhaps from the servers to your location and back or even within components within that data center facility.
So, we have got firewalls; we have got all sorts of different technologies that are brought to better secure the network itself and the information that’s being transported across the network.
Geni – We have got to worry about viruses and malware too. So, we want all that on the network as well.
Joanie – You want to protect the network from all these types of things so that you know segmentation, keeping different pieces of customer delivery separate from each other so that they can’t be mixed.
Geni – So, one of the things that I have run into and from my own experience being in a hosted environment is this ability to have a break on my data impact your data. How does that happen when you are in a hosted environment?
Joanie – Well, that kind of gets down to the architecture. You know the network and security the way that the service provider sets it up. There are circumstances where there might be shared resources. You might have a server, and another customer might have a server, but those servers might be connected via a common network. What that does is it creates a path from one machine to another perhaps so if your machine gets involved with malware or something like that it can easily traverse the network and go involve other customers.
So, you know, having security that creates those barriers in segments that protects and contains each of those deliveries is really quite important. Then we have to take it down to the user level.
Geni – So, what happens there. So, I have a team of remote workers. How can I make sure that one of those rogue workers doesn’t bring something into my entire environment?
Joanie – You know quite a lot of that is just user education. You know, don’t click on the email, don’t go to the malicious website, all of those kinds of things. You know user behavior is something that’s very difficult to control. I think most IT managers have figured that one out, which is pretty much why we don’t give most end-users administrative access to the system.
You know that’s part of securing the overall system. If my account doesn’t have permission to do terrible things, then it’s less likely that even if I make a mistake and click on that bad email, it’s less likely that it’s going to actually get somewhere into the system and do some damage.
Geni – So, rights and permission are going to limit my ability to damage to everybody else.
Joanie – Absolutely!
Geni – Great!
So Joanie, let’s talk about disaster recovery. We had fires in our Napa Valley, and our whole office was set up for remote access. We had a server in the office, and the electricity was out, so our servers were dead. We had people working remotely from homes and no ability to do anything. So how come we prevent that when we go a hosted solution.
Joanie – We touched on that a little bit, earlier when we were talking about redundancy, and that’s really the case. So, for example, your server – you have a backup.
Geni – Ya! and all my data. (laughs)
Joanie – That’s great! That’s awesome. (laughs)
So, you need to go get another server. Now you need to reinstall all your software, you have to reconfigure all your software, create all your users, all of this other stuff, restore your data, and now you are potentially back into production. So, there is a quite enough a lot of recovery there that has to occur before you can get back into any level of real productivity.
So, when we talk about disaster recovery, we are talking about so much more than just the data backup. We are talking about replicating the entire infrastructure, taking basically a snapshot of your server, and replicating that to another location and having that equipment essentially on standby.
So sitting there, it’s warm, it’s ready, and it’s loaded. So that if something happens with the facility or with the piece of equipment, there is another redundant facility or piece of equipment ready to jump in and take over that workload and that’s really unique amongst a lot of the service providers.
Geni – If I am running my own little server, there is no way that I am going to have a second one somewhere else completely setup. I can’t do that on my own.
Joanie – No, that’s generally a fairly expensive value proposition for a lot of businesses, and for the service provider, it’s also a very expensive value proposition. But that level of disaster recovery is the thing that allows business continuity. It allows the organization to not only stay operational but fully operational and keeping their people productive.
Geni – 30 staff people unable to work for a week; I mean, that’s a huge cost. It was worth it for me to be able to fund it, but somebody should have an insurance policy basically.
Joanie – That’s exactly right. So, when you talk to your hosting provider and you ask them about their disaster recovery or not talking about – do they just have copy of my data in a different location? The real question is can they take workloads that a fully operational in a facility and nearly instantaneous (within a very very short time frame)? Can they re-deliver those workloads from other facilities and keep the customer operational? That’s really what we are talking about with disaster recovery.
Geni – That’s a very different story than what we think. We all think recovery with that backup tape that we can’t get access to.
Joanie – Yeah, sorry (laughs)
Hate to tell you that tape backup is not really the rest of the story.
Geni – So, now I know why you have bunny slippers, Joanie. Because you are staying up all night, having nightmares about what could happen if somebody is not covered. So, you have to have good slippers to pace the house, to try to deal with your anxiety. Is that it! (laughs)
Joanie – That’s it. (laughs)
Geni – That’s what it is. OK, so thanks a lot for spreading the nightmares to me now. (laughs)
So now we know what to look for in a hosting provider. We need to make sure that they have a proper data center, that they have got all the security handled for us and that we are positioned well for business continuity and disaster recovery.
So now, with that, you are all now prepared to get your own bunny slippers and get ready to take a ‘Journey to the Cloud.
In this video series “Journey To The Cloud,” our leadership board member Geni Whitehouse shared her key insights on cloud and all its critical aspects, including types and deployment models.
By going through these videos, you can now choose a cloud provider by analyzing it on some essential parameters like performance, security, and business continuity. Moreover, Geni’s interview with Joanie Mann also made us aware of the factors that every cloud provider should take into account before offering cloud hosting services.
If you have any further questions regarding cloud, you can get in touch with a Solutions Consultant anytime at 855-223-4887.
Subscribe to Ace Cloud Hosting on YouTube to receive even more awesome and informational videos.