The Virtual View

VMworld 2008 – Heading Toward a Virtual Data Center OS

VMworld Las Vegas was the biggest ever – 14,000 people. The Main Hall where they served meals was the size of about 3 football fields. There were food hall staff lined up on both sides of the entrance forming a human corridor to funnel people to their tables. As you walked down the 100 yard aisle to your table the staff on both sides would greet you, which was very strange. It felt like I was in a parade and I should wave to everyone.

Once I got over the vastness of the show and focused on content, I found the sessions were outstanding. It all started out with a great keynote presentation from Paul Maritz, President and CEO, VMware. He introduced VMware’s vision for VMware becoming a Virtual Data Center OS (VDC-OS) that allows customers to virtualize and centrally manage all your IT resources.

My take on this is that the Virtual Operating System will pool all CPU, memory, networking and storage resources together and allocate them to applications on-the-fly with high availability and fault tolerance across the entire IT infrastructure. The whole idea is to eliminate the guesswork and manual configuration of data center resources and have it all done automatically based on whatever you set as requirements or Quality of Service settings for your apps.

I attended many technical tracks related to this theme. The goal for VM scalability in the vCompute area is to support 8 vCPUs and 256GB of RAM per VM, 40 Gbps of network throughput, and 200,000 IOPS. Next generation of resource pools will scale to 64 nodes in a single cluster, 4096 processor cores, 64 TB of RAM, and 6 million IOPS, all running under DRS. Similarly, LeftHand announced new SANs that scale to 80 CPUs, 320GB of RAM, 160Gbps of bandwidth and 40GB of non-volatile cache, all virtualized and managed as a single storage pool..

In the vNetwork arena, VMware introduced Virtual Distributed Networking with the vNetwork Distributed Switch. The vNetwork Distributed Switch abstracts the configuration of virtual networking from the host level to an aggregate data center level, simplifying the setup and change of virtual networking.

The way I look at this new virtual networking technology is that bandwidth from the clients to the servers, and from the servers to the SAN, can be whatever your applications require. For example, bandwidth to your iSCSI SAN can be 40 Gbps if required. This clearly makes the iSCSI vs. Fibre Channel SAN protocol bandwidth debate a moot point, especially when you have a SAN on the backend that scales out as well.

Another really cool feature that’s a long time coming is Fault Tolerant VMs using vLockstep technology. This technology maintains VM/server state across 2 servers, so when one fails the backup server is in the exact same state and can continue on with no data loss or downtime associated with manual failover. (The catch is you need HV-compatible processors.)

What’s really slick is when you combine Fault Tolerant VMs with LeftHand’s SAN/iQ synchronous multi-site software. LeftHand’s Multi-Site SAN software allows users to define multiple SAN data instances across physical virtual servers and locations, all virtualized and seen by the VMs as single instance, local copy of data, regardless of where it physically resides. When combined with VM Fault Tolerance, your VMs remain connected to their same storage volumes regardless of which physical server or geographic location they failover to, with all server and storage state preserved, delivering true VM and SAN fault tolerance and HA.

The best part was seeing LeftHand’s VSA software running in the VMware hands-on labs. So many customers came to our booth to ask about it, because these are obviously technologies that fit one another perfectly. In my mind, Storage Virtualization is all about providing simplified storage resource provisioning and management with complete abstraction from the hardware. How many storage companies can claim that they’re not tied to their hardware and can run as virtual SAN? That’s what made the show really special for me.

We had an outstanding show with over 1,000 people visiting LeftHand's booth, and loads of fun at LeftHand's VIP Reception for Customers, Partners and Prospects at the TAO restaurant.

VMworld Vegas TAO VIP Event

Virtualization

With VMworld right around the corner and Microsoft Hyper-V launched, it’s a good time to revisit what virtualization is all about. From an IT infrastructure perspective, it’s technology that allows you to consolidate servers and storage islands into a centrally managed, shared server and storage infrastructure. The purpose is to reduce space and power requirements and greatly simplify management.

This message is very appealing, but what customers really want to know before they jump on the virtualization train is:

 
1. Will the reliability and performance of my applications and storage meet requirements after consolidation?
 
2. Will I still have high availability and data protection in the event of hardware faults, network link failures and power outages?
 
3. Will I have the required tools to manage and monitor my new, consolidated infrastructure?

The most popular response from our competitors seems to be “it depends.” (It’s amazing how complicated the industry makes it to get straight answers from vendors to these questions.)

At LeftHand we believe we are on the right track to deliver high performance, highly available, consolidated storage solutions for virtual server infrastructures. I’ve had discussions with many customers and those that spent the time to understand LeftHand’s unique capabilities in a virtual infrastructure were pleasantly surprised to see how we solved these problems.

• 1.  Will the reliability and performance of my applications and storage meet requirements after consolidation?

If you deploy a LeftHand storage solution, the answer to this is a resounding YES.

In a virtual server environment many virtual servers access the same shared storage simultaneously. LeftHand storage systems are designed to perform best under multiple concurrent workloads. LeftHand also delivers rapid, easy provisioning to support virtualized applications that are frequently set up and torn down.

As the number of VMs grow, LeftHand SANs can be grown seamlessly to meet the new capacity and performance requirements without disruptive hardware or software upgrades along the way. Add a storage node in 5 minutes and new drive capacity, storage processor and network connections come to life, increasing the entire SAN’s performance. LeftHand also includes advanced performance monitoring with real-time storage performance metrics for servers/VMs and their storage volumes.

• 2.  Will I still have high availability and data protection in the event of hardware faults, network link failures and power outages?

It’s not worth deploying any new storage technology if it will compromise the availability and data protection in your virtual server environment. LeftHand provides the same level of HA and business continuity in every SAN, including virtual SANs created with our Virtual SAN Appliance (VSA™). SAN/iQ software at its very inception was designed to deliver extreme fault tolerance and high availability through the creation of virtual volumes across clustered storage nodes.

SAN/iQ allows users to define multiple SAN data instances across physical virtual servers and locations, all virtualized and seen by the VMs as single instance, local copy of data regardless of where it physically resides. This allows VMs to remain connected to their same storage volumes regardless of which physical server or geographic location they move to, providing true SAN data mobility and HA to complement Virtual Machine mobility and HA.

• 3.  Will I have the required tools to manage and monitor my new, consolidated infrastructure?

A LeftHand SAN comes with all the enterprise-class tools you need to easily manage and monitor your storage network. Our award-winning, user-friendly software lets you manage everything from a single console. Capacity-free volume/snapshot cloning is also included for the management of virtual machine, application and O/S images.

Our customers tell us this every day: LeftHand SANs and virtual environments were made for each other. They like the fact that you can add capacity on the fly without having to manually re-layout or re-provisioning the new and existing storage. And that our storage systems have the intelligence to automatically re-layout the data and re-load balance the existing and new storage resources.

You don’t have to take my word for it. We will have customers presenting at VMworld, and hosting educational sessions. Stop by our booth, and we’ll give you a live demonstration of LeftHand SANs and their unique capabilities for virtualized IT infrastructures.

For more in-depth discussions, join us at the following sessions:

LAVCC Virtualization 101 track with LeftHand Customer Jorge Mata. Learn how Los Angeles Community College built a world class disaster recovery infrastructure. What they did, how they built it, how they funded it and plenty of tips and tricks on what you can do.

Session BC2477: Real World Best Practices for Site Recovery Manager. A LeftHand system expert will teach you how to implement VMware Site Recovery Manager to take DR to a new level.

Session PO2647: Architecting Remote and Branch Offices for HA and DR. Learn how to implement virtualization cost effectively in your remote and branch offices to enable HA and DR.