• Solutions
  • By Industry
    • Distribution
    • Financial Services
    • Gaming
    • Government
    • Healthcare
    • Manufacturing
    • Media & Entertainment
    • Travel, Transportation & Logistics
  • By IT Need
    • Proactively Prevent Downtime
    • Protect Data for Compliance
    • Achieve Disaster Resilience
    • Reduce CAPEX & OPEX
    • Provide HA in Remote Locations
  • Small & Medium Business Solutions
  • Windows Business Application Availability
    • BlackBerry HA
    • Exchange HA
    • SharePoint HA
    • SQL HA
    • XenApp HA
    • BlackBerry DR
    • Exchange DR
    • SharePoint DR
    • SQL DR
    • Virtual Server DR
• Products
  • Overview
  • everRun 2G
  • everRun VM for XenServer
  • everRun DR
  • everRun SplitSite
  • everRun HA & FT
  • everRun Advantage
  • Product FAQs
• Services
  • everRun Services Overview
  • Consulting Services
  • Training Services
    • Course Schedule
    • Course Registration
  • Assurance Program
  • Customer Portal
  • Requesting Technical Support
  • Technical Community Login
  • everRun Assist Web Support
• Partners
  • Partners Overview
  • Citrix Alliance
  • Microsoft Alliance
  • Why Resell Marathon
  • Partner Program FAQs
  • Find A Reseller
  • Partner Login
• Resources
  • Resource Center
  • White Papers
  • Webinar Recordings
  • ROI Calculator
  • Collateral
  • Flash Demos
  • Customer Endorsements
  • Success Stories
  • Product Reviews
  • Analyst Opinions
  • Exporting everRun Products
• News & Events
  • Latest News
  • Press Releases
  • In The News
  • Events & Webinars
• About Us
  • Overview
  • Why Marathon
  • Executive Team
  • Board of Directors
  • Awards
  • Careers
    • Current Openings
    • Benefits
  • Corporate Offices
  • Contact Us

Blog

Thanks again to everyone who joined us for last week’s webinar “Top 5 Low Cost Tips for Preventing Exchange Downtime” where Marathon’s availability experts reviewed their key tips for the prevention of downtime, including:

1. Reduce human error with process
2. Document your infrastructure
3. Remove single points of failure
4. Don’t forget to test
5. Understand your requirements

There’s a lot of great information in this 40-minute webinar, so be sure to check it out. We’ve summarized the Q&A portion for the webinar below.
 

Q: What type of storage does everRun support?
everRun supports any type of storage that you have. The most common storage configuration we see is local disk drives for the servers themselves. That would have the same amount of data protection as even a SAN would, and in some cases would be even better protection, because you have total redundancy from both servers and everRun is protecting that as if it’s a single storage device. You could also have iSCSI connected storage, or any kind of SAN storage that you wanted to have. Again, everRun supports any type of storage.

Q: Which versions of Exchange do you support?
The beauty of the everRun architecture is that is can support pretty much any Windows-based application. Exchange 2003, Exchange 2007 or Exchange 2010 – everRun supports them all. Some other high availability solutions require specialized scripting to support applications, but everRun does not require this. Also, with solutions like clusters, sometimes you have to buy the higher-end more expensive “enterprise” versions of the application software to support that configuration, but with everRun, we can provide complete protection for the standard versions of Windows and Exchange Server or any other application.

Q: What is the load on the systems when using everRun?
The good news here is that there is very little overhead associated with everRun – about 5% to make things run redundantly. That’s a very small performance price to pay to get such a high level of protection for Exchange.

Q: How does everRun handle the mirroring of data that’s loaded in memory?
There are a couple of ways that is done. Since the application is actually running both severs simultaneously, that means the memory is being replicated on both servers simultaneously as well. Keep in mind that as the applications execute, they are storing to storage, and because of the redundancy built into the everRun solution, that data is being written from memory down under the storage element redundantly as well.

Q: Is it possible to run servers in two different locations?
Absolutely. In the slide where I showed the everRun architecture with the two servers, you can take those two servers and separate them geographically. They could be in different rooms in the same building, different buildings on the same campus, or even separated further, by about 100 miles, depending on the bandwidth and latency of your connection. We call this our SplitSite configuration.

Q: How is this different from a cluster solution?
The major difference of everRun vs. a cluster solution is that we are doing operations on two servers simultaneously. The application is actually running in tandem on both of these servers. With a cluster solution, you’re running your application on one server, while the other server stands by and waits for a failure to occur. That means that with a cluster solution, when the first server fails, the cluster then has to do something to start up the application on the second server and then continue from that point. But that means downtime, data loss, and loss of connectivity. With everRun, that doesn’t happen. Because the other server is already doing the same thing, there is no downtime and no data loss, because there is no “recovery” – even when there is a failure.

Q: So are both servers “hot” in an everRun configuration?
Yes – that’s exactly right. Both servers are active and run simultaneously, unlike a cluster. So with everRun, you could have a failure of a component on one server and then another type of failure on the second server and still be operational. With a cluster, this scenario is not possible. If you have failures on both systems at the same time with a cluster, then you are down.

Q: Does everRun require dedicated servers just for Exchange?
No – everRun protected servers do not need to be dedicated to one specific application. You can run multiple applications on this pair of servers, and even chose which ones you do or don’t want to protect with everRun. This is good for small businesses, who want or need to consolidate several applications on to fewer servers.

For more information about protecting Exchange from downtime, be sure to check out our white paper "Six Secrets to 24x7 Exchange Availability."
 

Earlier this week, we hosted a webinar on the topic of “How to Cut Risks and Costs with a Downtime Analysis & Action Plan.” We know from our experience in application availability that many companies avoid these types of assessments – they either don't know where to start or decide that they don’t have the time or experience to conduct an assessment, so they just live with the unknowns and hope that nothing bad happens. (We’ve seen the consequences of downtime at many companies and don’t recommend this method!)

Our VP of Services & Support, Beth Shea, explored this topic in detail and provided a simple framework that companies can use today to uncover their risks and put measures in place to minimize the impact of downtime. To learn more, be sure to watch the 30-minute webinar. You can also check out the Q&A session from the webinar, summarized below.

Q: When looking at the impact of downtime, it is just unplanned downtime, or should you include planned downtime as well?
You absolutely need to plan for both planned and unplanned downtime, as there’s a real cost and business impact to both. They both need to be included in your impact assessment.

Q: What about branch offices – should they be included in a downtime assessment?
According to Forrester Research, about 20% of a company’s business is tied up in branch and remote offices, and IT needs to include these offices in any assessment that they are conducting. You shouldn’t overlook these offices when putting together your downtime and business impact assessments. They have to be factored in.

Q: How often should I conduct a business impact and risk assessment?
What we’ve found with our customers is that conducting an annual assessment is sufficient, or in some cases, twice a year, depending on the type of business. You can then use these as your benchmark going forward to determine the success of the initiative and ensure that you have the key metrics to report to your management team.

Q: How do you determine when to use local high availability vs. a disaster recovery solution?
Fault tolerance, high availability, disaster recovery - all of these different terms can be confusing and they can have different meanings to different people. The way we think of this is that when you’re implementing high availability or fault tolerance this is to ensure that locally you are protected against the everyday, nuisance failures that cause downtime. If you lose a fan or a drive for example, you would automatically route to another server within the same building or local area. Disaster recovery solutions are really for recovery from catastrophes (fire, flood) or other events where you need to failover to a much more distant location. You don’t want to use this type of solution for everyday failures, as it can be very time consuming to failover and failback, and you can potentially lose some data. For local protection, you want high availability/fault tolerant solutions.

Q: What about hosted applications like salesforce.com, how do I account for those in this type of assessment?
In today’s world, so many applications are offered as Software-as-a-Service (SaaS) or sometimes called hosted applications, where they are no longer hosted at your site. However, they are still important to your business and need to be included as part of your overall assessment. Our approach is to conduct the assessment for your SaaS applications as if all they were onsite. Then use your tiered analysis and make sure that your SaaS vendor is meeting your availability requirements for that application, and that they have the necessary protections in place to protect that application to the same level that you would protect if it were in-house.

Q: Does Marathon offer any services to conduct this type of assessment?
Yes – this is a service that we provide for our customers. Most customers are very satisfied with the service, because it usually has an immediate ROI for their business. If you are interested in this type of service, please feel free to us at 978-489-1100.

Q: Does Marathon have any templates available to build a framework for this type of assessment?
Absolutely. From our 16+ years of working with customers on the assessment and prevention of downtime, we’ve put together an extensive list of questions to ask about the business risks and impact of downtime. Please feel free to contact us if you would like more information.

Q: How do you measure or put a price on the intangible impacts of downtime?
This can be tough to nail down, but what we recommend is developing some basic estimates. This isn’t meant to be an exact number, what we are really trying to achieve here is to prioritize applications, put them into the tiers that we discussed and make sure that you are putting the right amount of resources against the right applications. From a productivity perspective, one metric you could use is to look at the cost of employee salaries and how much it would cost in salary costs to have employees not be able to work for a certain amount of time. This is just one example.

Q: Does everRun handle quick switch over to back up site if the main site goes down?
Yes, within seconds.

Q: What are the requirements for the backup site?
The machines at the backup site are in the same pool as the primary site, so the backup machines must meet the requirements to be in the same pool as the primary site machines.

Q: How about regular data sync between main site and backup site?
Since the primary and backup site are running in lockstep mode, the application and the data are always in sync between the primary and backup sites.

Frank Ohlhorst, former Executive Technical Editor for eWeek and award-winning IT expert, was our expert guest speaker this week for the webinar, “Cut Your DR Costs and Get Better Data Protection.” During his presentation, Frank reviewed why he believes that now is the time to rethink traditional approaches to disaster recovery. He explained why the total cost of ownership for disaster recovery solutions is on the rise, and why changing data protection dynamics are making it more economical to focus your time and budget on the prevention of downtime and data loss, rather than recovery.

Below is the summary of the audience questions from the Q&A portion of the webinar.

Q: You talked about how HA can give you a geographic advantage. What do you mean by that?
Frank Ohlhorst: High availability systems are designed to work with multiple servers and there’s no reason why you can’t have those servers located hundreds or thousands of miles apart. You get a geographic advantage because your data centers is in multiple places and regional areas, so if a weather-related or other event occurs, let’s say a blizzard up north with a power outage, your data center down south can pick up the slack without kicking users off the system. The same can be said about a data center located in an area with hurricanes or other natural disasters. The geographic separation gives you added protection.
When high availability is paired with load balancing, it helps to locate the data resources closer to where the users are requesting them. Let’s say you have users in Utah, it’s better performance-wise to have them talk to the data center in Nevada rather than Virginia. It helps on that level also. HA solutions also have the tools for monitoring what is going on with your users and network, to help you plan out how you should assign users to specific data centers for the most efficiency.

Q: I understand how high availability can handle unplanned downtime, but what about planned downtime? Can it help there as well?
Frank Ohlhorst: Yes, the idea there is being as you have multiple active systems to meet the user’s needs, you can take one of those systems down for maintenance and have the users serviced by the active machines while you make the updates and improvements. Then when you are done, just resynchronize with the other systems, move the users over to those systems and update the rest of the servers.
Another great benefit of this is for testing upgrades and changes. So take one system offline and test your upgrades to see if they work properly before you return that system to production.

Q: If I have an HA solution in place, is back-up still necessary?
Frank Ohlhorst: 99% of the time the answer to that question is yes. It depends on what your corporate needs are. There are certain situations where HA might not deal with your catastrophe. Those are usually software-damaging events, like a virus infection, that winds up getting replicated across the system. Of course, that should really be part of your security planning to prevent events like that from even happening. With today’s security technologies, it’s pretty easy to prevent that. But if you did ever have one of those events, you do need something to roll-back to, and that’s where the back-up comes in to play. Ideally though, you should be preventing that type of event, because you also have the potential to lose active data if that happens. When it comes to compliance or auditing, you have to restore data relevant to that time period to meet the needs of e-discovery, compliance, accounting audits and other similar requirements. So you can’t just say, “I have HA in place, so I don’t need to back-up.”

Q: What about data de-duplication technologies, don’t they help solve this problem of managing large volumes of data?
Frank Ohlhorst: They reduce the data footprint for sure, but what we’re talking about here is availability of the data. They can certainly reduce the size of your data footprint, you can use de-dup to speed up backups. At the end of the day though, if the system or application is not accessible to the user, then it’s not available and you haven’t met your objectives. It’s a simple matter of business logic that data de-duplication can improve performance and reduce the size of the footprint, but it doesn’t solve the problem of providing access to users during catastrophic events.

Q: Do you see continuous availability and high availability as the same, and if so, how do you differentiate between the two and the costs?
Frank Ohlhorst: There was a time when those technologies were very, very different. That was way back when we relied on expensive hardware-based solutions or appliances that provided continuous availability. High availability at that time was thought of as a method to switch from one server to another using a manual process in the case of an emergency.

High Availability technology has evolved significantly since then. Now, the two are really one in the same from a planning and software point of view. Today’s HA solutions eliminate that step of manual switchover. What you see with the vendors today is automatic HA technology that really delivers continuous availability. And the cost gap today is pretty much zero, since the technology for continuous availability and high availability has evolved to be almost one in the same.

Q: With an SRDF/S-type solution, how can we get around the fact that being geographically more separated to mitigate regional disruptions can mean slower primary system response times due to the need to remain synchronous?
Frank Ohlhorst: Let’s look at this first from the ideology of what we’re trying to do which is business continuity. So, if you encounter a situation when you lose connectivity to a system and it’s still available at another location, then you’ve met the goal there of providing continuity. And you’re in much better shape than you would be at that point if you had a disaster recovery solution instead of a business continuity solution.

The question you have to ask yourself at that point in time is: Is reduced performance better than no performance at all? For most businesses, the answer is yes. For others, if the performance lag is significant enough it can impact business. In those cases, you’ll have to work out a way to develop geographically dispersed sites can that can provide enough performance to the user sets that need access to the data. You also need to make sure that your connectivity has enough bandwidth to support your BC/HA solutions, which means the ability to replicate the data in real time across the wire. You might have to invest in larger pipes for better connectivity to support that. But again, that depends on your particular business and your needs. There is no one correct answer to this question, but the good news is that there are several solutions today that can help you solve this problem and meet the levels of availability that you need for your business.

Earlier this week, we had a great webinar featuring guest speaker Greg Cullen, Sr. Director of Technology at Marathon. Greg provided tips and advice that government agencies can use to ensure optimal availability for their critical applications at minimal cost. He also reviewed  three different government agencies that protected their data and applications from downtime with Marathon’s everRun software.

These included the Brookline Police Department, which kept their 911 system available – even during a hardware failure on Christmas day when no IT staff were on duty; the City of Santa Rosa utilities department, that protects its water treatment facility against outages from earthquakes, power outages and other natural disasters; and a county court system in California that protects its virtualized Exchange server and other paperwork processing applications from downtime.

The transcript of the Q&A session with Greg from the live webinar is below.
 

Q: Can you deploy the two everRun servers in different locations or do they have to be in the same location?
You absolutely can. You can deploy the two servers in the same room, or different rooms in the same building and even in two separate buildings. We call that configuration SplitSite. That’s one way to get disaster tolerance in the event that you have a site-wide outage, rather than having them in the same room or building.

Q: Is there a limit to how far apart the servers can be?
In general, they can be separated by as much as 100 miles, although it really depends on your bandwidth and latency on the connection between the two sites.

Q: What are the hardware requirements for everRun?
You’ll need Intel-based servers that have a moderate amount of memory, or as much as the application requires. And you’ll need enough networks to do the production side of it as well as maintaining the redundancy between the severs. Generally speaking, having four network adaptors in each server and somewhere in the order of 100GB of disk drive is sufficient.

Q: How does the everRun software compare with a clustering solution?
One of the biggest difference between everRun and clustering is that everRun is a single image for the application. Instead of installing and managing two instances of your application like you do with clustering, everRun is just a single image to install and manage. Changes happen on both servers simultaneously through that single image.

Also, everRun software does not require cluster-aware applications. everRun is application agnostic, and can support almost any Windows application. And one more thing, with most clustering solutions, you also need to have a shared storage container that both servers are connected to. everRun can support that model as well, but doesn’t require it like clusters do. In fact, to remove single points of failure, it’s much better to have local storage connected to each of the servers and everRun will manage that storage as a mirrored device.

Q: I’m confused by your use of DR. Can you define what you mean by disaster recovery?
We’ve found that everyone has a different definition of what they mean by disaster recovery. At a very high-level, we see disaster recovery as the need to protect your data. By comparison, we see high availability as the need to protect your application, data and network connectivity. DR means you’re trying to copy your changed data off site to protect it in the event of a true disaster. After the “disaster” is over, you then need to bring that data back to the primary site, or configure an alternate server to use the data in the DR site.

Q: Does everRun work with Siemens building security systems?
Yes, we have been working for several years with building automation and security companies including Johnson Controls, Tyco, Andover Controls, Siemens and many others. As long as the building system runs in Windows Server 2003 or 2008, we can provide availability for it with no custom scripts or custom coding. We have many deployments of everRun protecting these building security systems around the world.

Q: Does everRun work with e911 systems?
Yes – absolutely. Generally speaking, everRun is application agnostic and can work with almost any Windows application. We have many solutions out there where these emergency 911 centers are protected by everRun so that if there is some type of disaster, these systems continue to run.

Q: Is everRun available on a GSA schedule?
Yes, through our channel partners. Contact your Marathon account representative or call 978.489.1100 for specific partner information.

Q: How does everRun differ from data replication solutions?
A lot of times when people look at availability, they simply try to replicate the data. There’s a big issue with that though. That’s only one part of what you need to recover in the event of a failure. everRun not only replicates the data, but also keeps a redundant set of your application environment and network connectivity, and everything else that is required for the application to not see any failures at all, or to recover very quickly in the advent of certain types of failures.
 

Earlier this week, Craig Resnick, research analyst from ARC Advisory Group, joined us to discuss Best Practices for Preventing Downtime in Automation Systems. Craig's presentation was very well-recieved, with several attendees commenting on the high quality of the information Craig provided. If you haven't had a chance to see it yet, the on-demand recording is here and the recap of the Q&A from the webinar is below.

Q: Has the hierarchy at manufacturers changed where the groups that mange these different domains have converged, or are they still separate?
Craig Resnick, ARC Advisory Group: Over the last five years, we’ve seen the convergence of IT with the automation and operations groups. Five years ago we used to joke about the “civil wars” between these groups. IT used to poke fun at the factory floor about the age of the equipment, which can be 10, 20 or even 30 years old in some cases. The Factor Floor used to poke fun at IT because, as they put it, IT didn’t understand what “real-time” means. We’re finding now that there are many initiatives between these groups to converge different processes at different levels. This is an ongoing process that will take a while, but from what we’ve seen, once the convergence is made, it usually has very positive results for the business.

Q: Is everRun tested and approved by Siemens, Rockwell, etc.?
Yes. everRun works with a number of different automation systems and applications from Rockwell, Siemens, Johnson Controls, Dematic, Wonderware and many others. We’ve done qualification and certification testing with many vendors in the automation space. Because of the way that everRun is designed, it is almost transparent to the application, so we really can work with most vendors and have a very quick validation/certification process.

Q: Will a TCP connection from a SQL client to a SQL server be maintained through a failover?
At Marathon, we take a different approach to application availability. It’s not about failover and recovery, it’s about keeping systems up and running, even during a failure, with no impact to the users or the data. Failover isn’t something that we really do. We can actually maintain those connections, even with a failure, at all times if that’s what you need. We can maintain all connectivity, transparent to the user and the IP connections, and keep the system states intact.

Q: Does everRun work in both physical and virtual environments?
Yes, everRun works in both physical and virtual environments. We can protect both single and multiple workloads.

Q: What is the typical integration period to get everRun up and running at a site?
A typical engagement is about 2-3 days. The software itself installs very quickly and then after that there is the deployment and migration of applications, testing and training. WE provide these services through our everRun ONE program.

Q: What is the typical overhead of everRun?
That will vary based on the application. Anywhere from 5-15% depending on the characteristics of the applications – storage intensive, I/O intensive, etc. But 5-15% is a typical estimate.

Q: Are the partnerships validated in both physical and virtual environments? We use the Dematic voice picking application.
We do support Dematic applications in both physical and virtual environments. Some of our vendors have only tested physical, some virtual and some both. Our technology is very similar for both physical and virtual, and in most cases will work with most applications in both. If you have a specific application that you would like to check on, just give us a call.

Next week Craig Resnick, research director and automation expert at ARC Advisory Group will be the guest speaker for our webinar "Best Practices for Preventing Downtime in Automation Systems."  We recently sat down with Craig to discuss some of the recent trends in the manufacturing and automation industries.

Q: What are some of the newer trends that you are seeing in the automation space?

Craig Resnick: A primary trend that we see at ARC is the convergence of automation and IT systems. Nearly every manufacturing company uses a variety of plant automation and enterprise IT systems to manage its operations. Plant floor systems, such as distributed control systems (DCS), programmable automation and logic controllers (PACs/PLCs), and a wide range of plant floor applications provide a wealth of real-time information regarding productivity, efficiency, equipment health, capability, and quality. Business systems, in turn, provide information on raw material costs, product orders and inventories, manufacturing resources, production schedules, etc. This wide range of information often remains isolated in systems such as manufacturing execution systems (MES), laboratory systems, maintenance systems, scheduling systems, enterprise resource planning (ERP) systems, supply chain management (SCM) systems, and customer relationship management (CRM) systems. Decisions based on data from any one of these system will always be less than optimal because, without the corresponding information from the other systems, the information will be incomplete.

To close this gap between automation and IT systems, and to address the trend of the plant floor becoming more IT-centric, ARC has defined a new space, defined as Collaborative Production Systems. These new systems consist of platforms in which the controls layer domains of process, logic, motion, building automation, and power control systems converge with the information layer domains of production management and MES systems. These converged systems enable, for example, the required data and information to be directly tied into applications such as corporate reporting and manufacturing compliance. Collaborative Production Systems will become the industrial blade server that provides full monitoring and control of the enterprise, from the office to the plant floor, sharing that information with the supply chain to, for example, procure materials and resources and purchase or sell power at the optimal times and prices from the smart grid, while providing full financial metrics and KPIs to ERP systems to maximize profitability.

Q: Now that corporate reporting and systems are heavily tied into the “factory floor”, how is that changing the need for system availability and data protection?

Craig Resnick: The need for system availability and data protection continues to expand, driven by a combination of issues ranging from global competition to regulatory requirements. Process safety and critical control are primarily focused on system availability and process uptime. As a specific example, take the Pharmaceutical industry, where data and batch information can never be lost or interrupted. System availability and data protection needs are also forcing E-records regulations to evolve across the globe. In the US, this includes 21 CFR Part 11, as well as the FDA’s Good Manufacturing Practice (GMP) and Process Analytical Technology (PAT) initiatives. In Europe, this includes Annex 11 of the EU GMPs, electronic Signatures Directive 1999/93/EC, and Data Protection Directive 95/46/EC. The European Data Protection Directive requires even more protection on data than the current FDA regulations and extends this requirement to clinical trials patients, as all clinical trials data requires maximum protection to remain compliant with regulations.

Unscheduled downtime is expensive. It often impacts production’s ability to meet its schedule and may cause missed customer commitments. Unplanned downtime, which also includes unexpected stoppages resulting from equipment failure, operator error, or nuisance trips, is the nemesis of all manufacturers. Statistics on the impact of unplanned downtime on plant operations show that it accounts for 2 to 5 percent of production lost in, for example, the petrochemical industry. Unscheduled downtime is also costly in terms of equipment damage, environmental harm, and worker safety. The cost of downtime is reflected in a primary key performance indicator (KPI) used by manufacturers known as Dynamic Overall Equipment Effectiveness (OEE), which helps determine the real-time impact of the performance of any individual process or piece of equipment on the overall efficiency of the plant. Unscheduled downtime is a primary factor that significantly lowers Dynamic OEE, which translates to the manufacturer decreasing both its efficiency and profitability.

Q: What are some of the basic steps that companies can implement to ensure the availability of their systems?

Craig Resnick: The first step that companies can implement to ensure the availability of their systems is to maximize their operator’s effectiveness in the control room, which is essential to minimize the risks of accidents, eliminate unscheduled downtime, and maximize production quality. The global process industry loses $20 billion, or five percent of annual production, due to unscheduled downtime and poor quality. ARC estimates that almost 80 percent of these losses are preventable and 40 percent of those preventable losses are primarily the result of human or operator error. Maximizing operator effectiveness requires automating as many functions as technology will allow, as well as reducing complexity wherever possible. For example there are still many plants where operators monitor the processes and collect data manually or semi-automatically using chart recorders. This process is both tedious and error prone, and does not provide appropriate process insight or instill a sense of ownership among the control room operators.

The Abnormal Situation Management Consortium (ASM) points out that most incidences occur from multiple modes of failure. Preventable human error is a contributing factor to these losses, but is hardly the only cause. Preventing abnormal situations requires a multilayered multi-discipline approach focused on maximizing production throughput, efficiency and quality while minimizing lost production time and preventing damage to assets and endangerment to personnel. This approach requires deploying collaborative production systems designed and implemented to be able to deliver high levels of availability and fault-tolerance expected from any other mission critical industrial system. This typically requires effective data backup mechanisms, redundant controllers for critical applications, plus industrial grade software. Manufacturers are also deploying more fault tolerant server technology to ensure continuous availability of these mission critical applications; the continuous flow of vital products to the market; and the avoidance of the potentially negative financial, social, or environmental impact that operating without high availability fault-tolerant systems might bring.

To learn more about preventing downtime in your automation applications, be sure to attend next week's webinar where Craig will provide expert info on steps for reducing the human error that leads to downtime, how to protect your hardware, storage and networks for complete availability coverage, and how to protect against a complete site failure. You can register here.
 

Marathon’s Sr. Director of Products, Michael Bilancieri, recently answered some questions for Paul Mah of ITBusinessEdge.com regarding disaster recovery planning for small & medium businesses. A few of Michael’s answers are highlighted below. For the complete Q&A with Paul Mah, see the article here.

Mah: Any tips to help SMBs with constrained budgets get management’s approval to implement a DR program?
Bilancieri: This may be the most important part of the process. Without support from the senior management team, any DR plan will be hard to get off the ground. The key takeaway here is to translate the technical language into business terms.

Since DR is not primarily about the technology (it is about the business value), it is important to clearly express what downtime means in terms of revenue loss. By creating a chart, organized by each application, it is easy to clearly articulate how much revenue is lost across each application for a certain amount of time.

Mah: What are the best criteria for determining an optimal disaster recovery plan?
Bilancieri: First, you have to identify what it is you need to accomplish. This includes defining the recovery time objectives (RTO), which is the amount of time applications can be unavailable and recovery point objectives (RPO), which is the amount of data that can be lost when a recovery is required.

Keep in mind that these values will likely vary for each of your different applications. Implementing incorrect or incomplete solutions will result in wasted time and resources. Check with your users and clients to determine their requirements and any service level agreements that must be met.

Mah: Once you determine exactly what your needs, how do you select a plan?
Bilancieri: DO YOUR HOMEWORK! Seriously, there are so many different products that claim to be “DR” solutions, all approaching the problem from different angles, it can be very confusing to determine what actually does the job you are looking for it to perform. As you research different products to implement as part of your DR plan, be sure to ask specifically what their product does (copies just the data, takes data snapshots, captures complete images of the full system, etc.) and don’t be afraid to ask probing questions.

Many vendors make the same claims using the same terms but actually deliver very different results. If you are going to test these solutions in-house, which is recommended, try to do the test under similar conditions as your production environment, with similar system and application loads. Oftentimes, something works well in a test environment [where there is] no real processing happening, [but] fails to function adequately once deployed in the live production environment.

Mah: What would a DR plan look like for a company that may face natural disasters such as hurricanes and flooding?
Bilancieri: Since hurricanes and floods can cause severe damage that can result in long-term outages, it would be wise to implement a solution that protects your systems between locations that could not be affected by the same disaster. Ensure that the backup, or DR, site is planned for a location that can be readily accessible by your users and clients should the primary location be destroyed or otherwise inaccessible.

Marathon has a customer based in Georgia, The Sullivan Group, which implemented a disaster recovery plan just for this reason. The team decided to virtualize its data center with Citrix XenServer and implement Marathon's everRun VM solution to provide redundant virtual machines and synchronized mirroring of the entire system including network, applications and data. The Sullivan Group has a small IT staff but needs to be continuously available for their clients, so they needed a solution that was fully automated and offered simply implementation.

Their first step was to identify what their customers’ needs were - and they decided that they needed continuous protection. Second, the team determined exactly what they could afford, and the ROI they would see from implementing DR software. They already knew that they would constantly face the threat of storms, and that they needed their data to be backed up in a remote location. Finally, they determined exactly what solution their IT staff could support and decided exactly which business applications needed to be fully available.
 

Jim Welch, Marathon's President and CEO is featured in this week's Worcester Business Journal in the "Shop Talk" column. Here's an excerpt from the interview:

WBJ: How has this crazy economy impacted Marathon Technologies?

Jim Welch: I think, prior to me joining {Marathon}, sales cycles were getting longer and people were being more careful about what they were buying. However, the nice part of our core business is, if you need it, you need it. If you’re putting in an application that has to run and be reliable, you don’t have a choice. It’s part of the infrastructure that keeps your business running. So, from that perspective, we’ve weathered the storm fairly well and I think as you look forward, now that that general IT spending is starting to ease up a little bit, we’re going to see a lift.

WBJ: Who are your customers?

Jim Welch: Typically, our best customers are ones that need their systems to never go down. But if you look at how things are changing over the last couple of years, we see a drive by IT shops to reduce costs by consolidating infrastructure. The way they consolidate is by putting more applications on fewer servers which stacks up their application risk, if you will, so if that one physical hardware fails, not one application fails but now three, five or eight could fail. So, in those environments we’re becoming more important as part of that infrastructure so they can rely on fewer servers.

Be sure to check out the rest of the interview, including the WBJ's interesting photo style on the Worcester Business Journal website.

In case you missed it, we announced this week that Jim Welch, former IBM VP and GM, has joined Marathon as our new Presdient and CEO. To learn more about Jim and the future direction of the company under Jim’s leadership, be sure to check out some of the recent press coverage:

Xconomy, "Marathon Hopes to Go the Distance with New CEO"

Mass High Tech, "IBM vet Jim Welch named CEO at Marathon Technologies"

ZDnet, "Conversation with Marathon Technologies"

Worcester Business Journal, "Marathon Hires Former IBM Exec As CEO"

Douglas Brown of www.dabcc.com recently interviewed Michael Bilancieri, Senior Director of Products and Tom Reed, Senior Systems Engineer. Michael, Tom, and Doug discuss the Marathon everRun high availability solution, what's new, how it works, how it adds value to Citrix XenServer and Microsoft Hyper-V, and much more.

Listen the Show

Can virtualization move beyond test and developmental in most companies and become a key component of the core enterprise IT infrastructure? Jerry Melnick discusses the necessity of high availability in a virtual environment and how it can be achieved with Mike Vizard of eWeek.

If you’re interested in using virtualization for essential applications like Exchange but haven’t yet because you’re worried about keeping them available, we encourage you to listen to this podcast and learn about the proliferation of different virtual machines, why there are different types available, and how Marathon’s HA solution can make virtualization of applications like Exchange and SharePoint a reality.

We’re constantly getting questions regarding the what, when, where, why, and how Marathon can help CIOs expand the use of virtualization and availability in their IT infrastructures. Jack Germain of LinuxInsider chatted with Marathon’s CTO Jerry Melnick recently and asked him some of these same questions, which we have excerpted:

LinuxInsider: Numerous industry reports show that the rush to adopt server virtualization is run mostly by very large corporations with a large bank of servers. What part of the market is your company targeting? Are you following the big money trail?

Jerry Melnick: Our product is designed to work like a standard application taken off the shelf and installed right out of the box. Existing availability systems can cause more problems than they solve. Our system is different. That is precisely what midsized companies need.

LI: What barriers does everRun VM take down that you expect to drive a company's need for what you offer?

Melnick: For one, there is a cost advantage over high-end proprietary products. On the other end, data replication solutions don't provide all the functions needed. Replication is managing data only. The comprehensiveness of our solution is what makes it unique. The application is unchanged in our environment. In addition, our solution does not need lots of setup.

LI: What are you thinking about beyond the accomplishments of everRun VM as the next virtualization enhancement?

Melnick: Now we have taken the process to the next leap. What I mean is we can now take two servers with hardware virtualized and build a Xen server pool and manage it as a single entity. We can build a virtual machine in Windows by installing out software and give it an interface of all running servers. Our next vision is to have one mechanism to do availability with multiple uses. It will take six months to one year. This will add new levels of choices to select plug-ins for desired results. All of this will be possible through one product.

LI: How do you see this vision changing the industry?

Melnick: We're breaking through limitations. We are in a position to expand use. Availability is something that everybody is talking about. We are providing ways of doing it.

For Germain’s complete article visit here.

Brian Ducharme from Virtual Strategy Magazine interviewed Simon Crosby at Citrix Synergy and got some great insights from him into the virtualization products at Citrix, their work with XenDesktop and XenServer, and how they work with partners (like us haha). Click on the video below to watch for yourself!

See additional insights from Synergy by David Marshall found here.

We’re elated with the responses we’ve been receiving over the launch of everRun VM. We had an opportunity to speak with some of the great media minds in the virtualization technology space and can’t thank them enough for their time and the positive response to the product launch. To date, we’ve seen everRun VM mentioned in 30+ publications online and offline and we hope to continue the conversation with everyone once the results of the initial beta test are available. As always, we like to recognize the people who spoke with us and took some time to share the everRun news with their audience. We encourage you to not only read the links below, but get to know each of these great minds below and follow them closely. Leave them a comment if you read something you like…trust us, they’ll appreciate it.

Virtualization for Everyone (Interview with our CTO, Jerry Melnick)
Tarry Singh
Virtualization for Everyone

Virtually Speaking
Dan Kusnetzky
http://blogs.zdnet.com/virtualization/?p=374

Network World
Laurianne Mclaughlin
http://www.networkworld.com/news/2008/032408-marathons-virtualization-tool-simplifies-disaster.html

For those who might have missed it, Tarry Singh interviewed Citrix’s CTO, Simon Crosby at VMWorld Europe last month. Check out the video below and see what Simon and the Citrix team have in store for the future (Be advised the video is a bit long so get comfy and grab a drink :) ). You can also read Tarry’s remarks on his blog here.

Over the last few weeks, many of our readers as well as customers have been asking us for more information about who we are and what we stand for. Because the virtualization market is such a broad waterfront, we find it necessary to make it as clear as possible how we complement the Citrix XenServer Enterprise environment e.g., creating a resilient disaster avoidance virtual environment that protects essential Windows Server applications from unexpected outages and unnecessary complexity :) .

Rather than answering every question here on the blog, check out the interview between Tarry Singh and our CTO, Jerry Melnick as they go deeper into our business goals and offerings.

For those of you who aren't able to attend iForum and have yet to see our fault tolerant virtual machines demo here is an interview with Virtual Strategy Magazine shot yesterday. Enjoy!

Recently, Gary had the opportunity to answer a few questions from Longworth Venture Partners, Paul Margolis, regarding the emergence of the high availability market and the benefits of virtualization. The Q&A session that was held can be found here. We encourage you to read through it and have a look at our take on virtualization coming to the high availability space. If you have any feedback, leave us a comment.