F5’s Kevin Hohenbrink: Data Replication Disaster Recover: Part 1

Upcoming Events

Your browser does not support JavaScript. This media can be viewed at http://www.podtech.net/home/2095/f5s-kevin-hohenbrink-data-replication-disaster-recover-part-1

F5's Kevin Hohenbrink: Data Replication Disaster Recover: Part 1

11:53 | Michael Johnson | Feb 12th, 2007 4:37pm |

Kevin Hohenbrink, product manager at F5 Networks, outlines the key points of data replication, recovery-point-objective (RPO) and recovery-time-objective (RTO), and their importance in a business continuity/disaster recovery plan. Hohenbrink is the optimization manager for the WANJet, F5’s appliace-based data compression and accelerator tool. This is the first of a two part interview. This is an F5 podcast.

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.

Tags: Kevin Hohenbrink, F5 Networks, WANJet

This entry was posted on Monday, February 12th, 2007 at 4:37 pm and is filed under PodTech, F5 Networks Incorporated, Corporate, Technology. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.