Following last week’s discussion, event attendees had additional questions that we didn’t get to answer even though we went ten minutes over. We wanted to continue the discussion here on our blog so we figured we would post the continuation of questions and answers for everyone to see. As we mentioned before, if you would like to view the presentation delivered last week by John Humphrey’s (IDC), Simon Crosby (Citrix) and Jerry Melnick (Marathon), download the presentation here.

Are there any performance limitations with everRun VM?

everRun VM supports any guest environment created by XenServer, including multi-CPU VM’s.

Effect of losing inter-server link?

As a best practice we recommend two Availability Links for redundancy. If one should be lost, we will continue to operate unaffected using the remaining one. If both are lost we will take action to prevent complete loss of the VM or SplitBrain.

How far apart can the two machines be – i.e. is there a propagation delay issue?

Host separation is a factor of network latency, which must be <10ms round trip. Current deployments have exceeded 100 miles.

In case of a disk failure, does everRun rebuild the disk from the good physical host to the bad one?

Correct. Recovery of storage is handled as a background task so as not to require downtime or otherwise impact the running VM and application.

When will level 3 of everRun VM be available?

Level-3, System-Level fault tolerance is scheduled for later this year.

What requirements are associated with the everRun Level 3 Protection? (Bandwidth, latency, etc.)

Network and configuration requirements are the same for level-2 and level-3 protection.

Is StorServer a similar or competitive product to everRun?

StorServer is a backup appliance, not a fault-tolerant availability solution, and addresses very different requirements. It would be more complimentary then competitive.

What virtual machines (VMware, Parallel, etc) are supported by Marathon?

Currently only Citrix XenServer, however future plans are to expand upon this.

Are there certain applications that are not suited for everRun, such as I/O or compute intensive apps? Home does DR configurations affect performance?

This is very dependent on the configuration of the server, the VM, the storage and all other components. Appropriate best practices should be followed to ensure optimal performance for all applications.

Can Marathon support physical to vm HA? Does Marathon’s product fully support FC/iSCSI SAN shared storage between protected physical and/or vm pairs? Does Marathon product support local site HA server pair with a third node at a remote site in the event of site failure? Does Marathon product have latency limitations?

Marathon offers solutions for physical and virtual servers. These solutions utilize the same proven fault tolerant technologies however are independent of each other. everRun VM supports any type of storage that is supported by XenServer. Fault tolerance is configured using two VM’s. However we will soon be releasing an asynchronous solution that will allow a third replicated system at a local or remote site. Because everRun VM is a synchronous solution there is a latency requirement of 10ms round-trip between hosts. Our asynchronous solution will not have any latency requirements.

What is the pricing of everRun VM?

everRun VM lists at $4500 when bundled with XenServer Enterprise, and $2000 if you already have XenServer.

Thanks for all of your interest and questions.

On Friday, CIO Magazine included us in its “10 Virtualization Companies to Watch in 2008.” Considering that the virtualization movement itself is something to watch in 2008, we couldn’t be happier. So, we know what you are all asking. What can we expect from Marathon in 2008? Well, that’s a topic for another post, but here is what you can expect from the blog in 2008.

1) Marathon Technologies “Are You Available Tonight?” blog will continue to keep you updated on all things Marathon and more. We’ll introduce a few guest bloggers now and again to keep things fresh and stimulate the minds of our readers.

2) We’ll link to Tarry Singh at least a few times because we think Tarry is insightful and we think our readers will like him to. (Read his post on the CIO Article, here)

3) We’ll avoid unnecessary “link love.” Although there are a lot of bloggers we enjoy reading, there is no need to link to them here, here, here and here.

4) Maybe, if you’re good, we’ll add a few pictures to really enhance the experience… but only if you’re good.

5) We’ll take requests… we mean for blog topics, please send your requests for Taylor Hicks to Ryan Seacrest.

So, that’s what you have to look forward to. It’s going to be a big year for the blog and we’re glad to have you aboard. Congratulations to the other 9 companies featured in the CIO article. We’ll be watching.