Part one here.

Transcript:
Host: Michael Johnson - PodTech
Guest: Kevin Hohenbrink - F5 Networks

Michael Johnson - PodTech
This is Part 2, in our discussion with F5 Networks’ Kevin Hohenbrink, talking about Data Replication, Disaster Recovery and Business Continuity and how F5 Networks’ WANJet product can help?

Now, one term that’s floating out there is business impact analysis. Now, how is this important to the overall Business Continuity and the Disaster Recovery Plan?

Kevin Hohenbrink - F5 Networks
Sure, an enterprise to undertake a business impact analysis to map their dependencies between critical business operations and the people resources applications and physical IT assets that they rely upon. In today’s environments, business process rely on multiple integrated applications, database and storage systems and so on. So, in order to restore a business process in the event of a disruption, you have to be certain replicating and coordinating the recovery of all these dependent applications. When you’re defining RTO and RPO application by application and selectively replicating data, it could mean, you’re only going to get a partial restoration of your business process.

Michael Johnson - PodTech
One of the other terms that’s out there, Kevin, it’s the term Consolidation. Now, how does this relate to protecting sites that you’ve got, say in remote areas? A lot of businesses right now have data centers, information applications and people information out in remote centers, especially as we’re expanding across greater and greater areas with globalization. How does Consolidation work?

Kevin Hohenbrink - F5 Networks
Sure, so data loss at remote sites is currently a huge risk exposure for most enterprise customers. Leveraging existing investment and existing data centers on recovery sites to offer Consolidation Backup Solutions from remote sites where centralized facility is critical. So, Consolidation will help ensure the backups to run regularly and very successfully. They’ve improved the manageability with their central administrators by providing visibility into remote site data center protection and potentially the ability to remotely configure and manage this protection, all with the goal of enabling their remote office to recover from the localized disaster.

Michael Johnson - PodTech
So, business maybe thinking about — maybe putting in some additional bandwidth to support those remote sites that we’ve been talking about or maybe improve some of the performances, some of their existing Data Replication Technologies or expand them as well. When is it important for a business with this kind of set up, to consider WAN Optimization and the products that are out there for it? What kinds of things should they look at and what’s an important consideration?

Kevin Hohenbrink - F5 Networks
Sure, so when you’re considering a WAN Optimization Solution, an important aspect of the WANJet appliance is that F5 has already taken the time and made the investment to test the interoperability of WANJet appliances with independent software vendors, storage vendors and storage network vendors like EMC and Double-Take. We’ve gone to the trouble of creating case studies in customer reference and it can provide customer references to prove out its capabilities and of course its intended benefits to our interested customers.

Michael Johnson - PodTech
Okay, so you have some folks that are looking at their networks. What kinds of things should they be considering in designing that network when they are looking at a Disaster Recovery Preparedness Plan?

Kevin Hohenbrink - F5 Networks
To achieve the desired RTO and your RPO with the greatest possible distance between the sites, enterprises are often going to design multi-site disaster recovery configurations to combine synchronous, asynchronous or even batch or schedule replication technologies. Site preparation is crucial because enterprises need locate their recovery site far enough away to escape the likeliest of local and regional threats, such as natural and man-made disasters. Synchronous Replication ensures the zero data loss because technically speaking, it does not return a right acknowledgment to their application until the data is being run to their recovery site. So, as a result, Synchronous Replication requires high-bandwidth, very low latency between the data centers in a metro area and the distance is no more than a 150kms apart.

Asynchronous on the other hand are typically deployed for long distances. With Asynchronous Replication the primary and secondary sites will be slightly out of sink and there’s some chance of data loss in the event of a disaster or a business disruption. However, applications are not forced to wait for the remote site to confirm a right acknowledgment before processing can continue and that’s the difference between those two.

Michael Johnson - PodTech
So, when companies are looking at the idea of data replication today, what are some main factors that are challenging folks and how can they be solved?

Kevin Hohenbrink - F5 Networks
Sure, so some of the key network challenges are, the expanding distance between DR sites. RPOs and RTOs are decreasing and a larger volume of data is needed to be replicated and of course TCP is becoming more common for all applications. This typically results in the WAN being seriously impeded from a performance perspective for it’s replication solution. So, how we’re going to resolve this? Well, you’re not going to ask the customer to replicate lost data, that’s not going to happen. Should the customer release more bandwidth? Well, an option, but it’s not a desirable choice, as this represents — as we talked about earlier, 20-30% of cost for data replication and it’s also recurring cost monthly. Accelerating a traffic — well, this is the most cost-effective method and this is what drives a solution like WANJet. Prioritizing the data replication traffic and guaranteeing bandwidth. This is what we talked about the quality of service capability of WANJet. This protects the traffic from WAN congestion or latency concerns.

Michael Johnson - PodTech
Can you tell me some of the factors, WANJets got operate in the real world too. So, what would be some factors that would affect the ability of a WANJet to accelerate that replication traffic?

Kevin Hohenbrink - F5 Networks
Sure, so the factors you are going to need to understand are the amount of repetition in data, even at the byte level, amount of compressible capability in the data. IE text is easily compressible, images are less, so all right. Another factor is the amount of different types of traffic competing for the same bandwidth. This requires WANJet to begin enforcing bandwidth guarantees which we can significantly improve performance of the important traffic at the expense of less important traffic. As I’d alluded to you before about the quality service capability.

Then the variability of data, when traffic is highly variable, congestion levels that would otherwise bring a replication process to halt, can now be prevented using bandwidth allocation because now we’re guaranteeing the bandwidth to that replication application.

Michael Johnson - PodTech
Okay, so tell me your company installs the WANJet appliance, what would be some measurable benefits that they could expect to see in their data replication strategy right away?

Kevin Hohenbrink - F5 Networks
Sure, so F5 has several customers who have deployed WANJet with industry leading storage replication vendors like EMC, Symmetric Solution and Double-Take. Some of the measurable benefits we’ve found were — doing our testing was the customers can meet their RPOs and RTOs without upgrading bandwidth or data replication solution infrastructure and they were able to accelerate replication traffic as much anywhere from five-ten times faster. They also were able to utilize 70% to 90% less bandwidth because we guaranteed the bandwidth and we prioritized for data replication applications over non-data replication traffic. So, you were able to guarantee the bandwidth for your data replication traffic because you can do the prioritization.

You’re also able to mitigate the total effective latency, affecting the replication process. We also enable them in their networks to adopt dynamically to data replication applications needs and congestion. We’re also it was kind of cool it was that we move more of the control of RAM resources into the hands of the storage team to depend on it and typically they have never had that kind of control or visibility, and then lastly, we were able to encrypt all traffic using SSL and of course this is a optional benefit on all WANJets.

Customers found that they can reduce the cost of meeting these RPOs by using a fraction of the bandwidth to replicate the same data, we were also able to provide a comprehensive view of WAN performance matrix and bottleneck so we could tell them who are the trouble spots were and then we were able to reduce the tangible and intangible cause to troubleshooting which sort of raises a big question for people. Back to the customers don’t have to incur the cost of bringing out the storage replication vendor on site.

Michael Johnson - PodTech
Well, Kevin you gave a lot of information here today. Is there a Website that folks can go to at F5 Networks or a particular place so they can go to find out little bit more about the WANJet Optimization Product.

Kevin Hohenbrink - F5 Networks
Sure, you’re going to go to F5’s new Website www.f5.com under products and you’ll search for WANJet and there is plenty of great white papers on the technology, white papers on our data replication, interoperability story with companies like Double-Take and EMC.

Michael Johnson - PodTech
Okay, so that’s www.f5.com, look under products it will we search for WANJet.Kevin Hohenbrink - F5 Networks
That’s correct

Michael Johnson - PodTech
Alright, Kevin Hohenbrink is the Product Manager for the optimization product WANJet from F5 Networks, it’s been a fascinating conversation Kevin, and hope to continue this as well as we explore more ideas about how to really make businesses and disaster recovery and threats to business and all of that easier through these types of products and it’s been great talking with you today on the Podcast.

Kevin Hohenbrink - F5 Networks
Well, thank you very much.

Copyright ©2006 PodTech.net. All rights reserved. Privacy policy

Tags: Kevin Hohenbrink, F5 Networks, disaster recovery, WANJet

Part two here.

Transcript:

Host: Michael Johnson – PodTech
Guest: Kevin Hohenbrink – F5 Networks

Michael Johnson – PodTech
This is Michael Johnson and we have on the line with us today Kevin Hohenbrink who is the optimization product manager for the product WANJet over at F5 Networks. So welcome to the podcast, Kevin.

Kevin Hohenbrink – F5 Networks
Well thank you.

Michael Johnson – PodTech
Now today we’re going to talk about some interesting things that have to do with how a business runs. We are going to talk about Disaster Recovery and Continuity. Now these are couple of terms that we hear are RTO and RPO and its importance to the Continuity Disaster Recovery Plan. Let us define what those terms are and explain how that works in this picture.

Kevin Hohenbrink – F5 Networks
So the two terms RTO and RPO, Recovery Time Objective and I’m going to use some sort of industry definitions that are pretty accepted by everybody. This is the maximum viable downtime after an outage for recovering systems, applications and functions. RTO provides the basis for developing cost effective recovery strategies and effectively getting the resources up and working again and implementing these recovery strategies during a disaster situation. Typically companies will measure in minutes to hours the downtime, their RTO and RPO. There was a recent study done by Forrester that talks about – in the event of a primary data center site failure, that 45% of North American respondents and 47% of European respondents recover in five hours or less. However, only a small percentage of those respondents, 8% North America and 7% European Theater could measure their recovery time in 120 minutes or less. Recovery point objective as per the industry standard definition, defines how current or fresh the data is after a disaster. Recovery point objective, the RPO, is really the earliest point in time which systems and data must be recovered after an outage. RPO typically defines maximum amount of data that the organization is willing to sacrifice after a disaster. And the zero RPO business continuity solution can survive a disaster without any loss of data and that typically tends to be very expensive. Another data point from a Forrester study — the same Forrester study, in the event of a primary data center site failure there was as much as 55% of North American respondents and 59% of European respondents. They would lose about five hours of data or less. 28% of the North American respondents and 27% of the European respondents could measure their data loss in less than 120 minutes. So when you put these two together RTO and RPO, they provide a measurable target for business continuity and Disaster Recovery Solutions. At any time you can improve the RTO and RPO, you got to increase your investment in networking and storage technologies as a result. The physical distance between your data center is typically and how well your applications tolerate network latency affect how close you’re going to get to zero RPO. This is why you should limit your RTO and RPO to whatever levels your organization can tolerate from a cost perspective.

Michael Johnson – PodTech
Well let’s talk a little bit talk about that cost and what’s the impact of the WAN on those recovery objectives when we talk about that distance and how far out your network goes?

Kevin Hohenbrink – F5 Networks
Okay. So a comprehensive Disaster Recovery Solution typically requires an investment, multiple hardened (ph) recovery sites, duplicating the IT assets such as your servers and your storage arrays, your networking equipment and you typically do it at all these sites and then having the replication software and the necessary bandwidth between these sites. Typically, the cost of bandwidth is often a significant component of the cost of Disaster Recovery Solutions that rely on data replication between these sites. There was a Forrester survey that said 25% of North American enterprises and 26% of the European enterprises reported that cost of bandwidth was representing between 20 and 30 percent of the total cost of data replication and again these are recurring monthly costs.

Michael Johnson – PodTech
Let’s talk a little bit more about that network — the idea of the transport network. How does that affect your recovery objectives in the long run?

Kevin Hohenbrink – F5 Networks
So the amount of bandwidth and the type of network transport selected whether it’s Wavelength Sonic Ethernet or IP is really the key to achieving desired recovery objectives. Limiting the impact and latency to the business applications and increasing the distance between the sites. WAN connectivity issues such as latency, reliability limit to service options and limited bandwidth may make significant impacts in improving recovery objectives and we’re going to talk more about this as we go through this, I’m sure.

Michael Johnson – PodTech
Let’s talk a little bit more about some of those characteristics and break that down a little bit more because I think it’s — we kind of went over it kind of quickly but I think they’re pretty important points.

Kevin Hohenbrink – F5 Networks
Sure. So the inherent characteristics you are going to find on a WAN network are latency, this is typically caused by limits to the speed of light over distance. You are going to have packet loss, caused by signal degradation over the network medium, over saturating network links, corrupting packets, rejected in transit or faulty network hardware. Network congestion, a big key point — excessive, lots of data on the network slows overall transmission speed kind of like too many cars on the freeway. Actual bandwidth is not the expected bandwidth often due to a combination of the factors listed above, whether its latency, packet loss or network congestion and of course last is expensive bandwidth. Large pipes can incur significant monthly costs. Unfortunately, such factors as these can often cripple a D R plan. WAN links are often subject to variable congestion caused by other application traffic, file transfers, even possibly other migration or recovery activities. This means your RPO and RTO that are met in minutes, can now be completely unobtainable the next minute due to congestion. So heavy latency do perhaps to extended distances can prevent meeting RPOs and RTOs irrespective of how much bandwidth is used. So adding more bandwidth doesn’t always solve all of your problems.

Michael Johnson – PodTech
Okay and now when people think about networks going down, sometimes they think okay, you know it is time we take a coffee break and you know, that’s kind of what it is but its — we’re talking a lot more stuff is on the table here besides this lost revenue, just things going down. What are some other key factors that are you know going to fuel that need to really improve your recovery capabilities of your business?

Kevin Hohenbrink – F5 Networks
Sure. So it’s also the cost associated with permanent customer loss and the ability of competitors to gain market share. So aside from the cost of downtime, additional drivers feeling the need to improve recovery capabilities are going to include increased risk fiduciary responsibilities to your shareholders, competitiveness in the market and of course regulatory like SOX and HIPAA are additional legal drivers. So based on recent events especially here in North America including terrorist attacks, blackouts, earthquakes, hurricanes wildfires and on and on, the perceived risk level is increasing among enterprises. So and additionally due to a significant number of corporate scandals that led to such government oversight and regulatory Sarbanes-Oxley, enterprises that operate in the business environment have increased fiduciary responsibility to their shareholders, partners and customers and even their own employees.

Michael Johnson – PodTech
Okay, now F5 is in the network business. Is there a WAN optimization appliance solution that you have?

Kevin Hohenbrink – F5 Networks
Absolutely. The WAN optimization solution from F5 is the WANJet product. F5’s WANJet is an appliance-based solution that uses compression, acceleration technologies to dramatically improve the speed of application traffic over the WAN. WANJet accelerates a wide variety of application traffic types including data replication which is the focus of this discussion, file transfer, email client/server applications and others. WANJet also has some unique features that enable bandwidth to be efficiently allocated amongst different applications, we call it our quality of service, and thereby ensuring that the most critical traffic receives the priority access to the valuable bandwidth. We buy the bandwidth in lot of cases specifically for a particular application, you want to make sure that that application gets its bandwidth.

Michael Johnson – PodTech
Okay, so can WAN acceleration appliances like this actually help in achieving that goal that we were talking about before, that zero RPO?

Kevin Hohenbrink – F5 Networks
Yes. WANJet appliances help to improve throughput. They also mitigate latency of existing networks through such techniques as this compression, our TDR data reduction and transport protocol acceleration. Often the cost of deploying a WANJet appliance at each end of the link is less expensive than the cost of increasing bandwidth. Typically, these appliances can be particularly helpful for enterprises that want to use replication or remote backup between sites, with limited bandwidth to the corporate data center.

Michael Johnson – PodTech
So how is WANJet going to work in that formula, to mitigate that latency?

Kevin Hohenbrink – F5 Networks
In situations where the WAN link, WAN is the bottleneck, WANJet can improve the performance of synchronous and asynchronous replication solutions which in turn can mitigate application latency and its performance impacts on applications.

Michael Johnson – PodTech
Okay. Now you also have this situation that, you know that enterprises that have some replications solutions in place between their data centers and other sites can the WANJet support replication of more data with existing bandwidth as opposed to say adding more, you know?

Kevin Hohenbrink – F5 Networks
Yeah, absolutely. Traditionally business applications like ERP, SCM, CRM as well as messaging and collaborative applications such as email — these are going to continue to grow steadily each year. These are often the very applications that are supportive with remote replication. The WANJet appliance can help enterprises support to continue replication of these applications with existing bandwidth. WANJet uses a QoS technique to guarantee and prioritize data replication over non-data replication applications as we just mentioned.

Michael Johnson – PodTech
Can it also enable the extending replication of other applications as well?

Kevin Hohenbrink – F5 Networks
Yeah. Do the cost replication — most enterprises are very selective about which applications they replicate and which ones they don’t. Usually they limit it to mission critical apps. Today it’s no longer a one-to-one relationship between a business process and application. Business processes now rely on multiple applications and to restore the entire process that means that they’re going to have to coordinate the recovery of the multiple apps. So customers are going to be very pleased to know that these apps that were once deemed only business critical as opposed to mission critical also require replication to another site.

Michael Johnson – PodTech
That was F5 networks Kevin Hohenbrink. This has been the first of a two-part series on data replication, disaster recovery and business continuity. Tune in next time right here on PodTech for Part two of our conversation with Kevin Hohenbrink.

Copyright ©2006 PodTech.net. All rights reserved. Privacy policy

Tags: Kevin Hohenbrink, F5 Networks, WANJet