Oct 9, 2013
When we shipped our Velocity Replication Acceleration software, or VRX, in February of this year, it created a new category of software to help storage administrators meet their recovery point objectives (RPOs) and backup windows. Before February, the only solution was full-blown WAN optimization and the myriad of deployment scenarios that came along with it.
WAN optimization typically lives at the edge of the network, near the WAN router, and uses some sort of network-based redirection. Network redirection is done with PBR, WCCP, VRRP, or just deployed in-line (bridge mode) without any redirection. All of these alphabet soup methods are network-based, require network expertise, and sometimes require network re-architecture. Does your head hurt yet?
With VRX, Silver Peak wanted to radically change and simplify the way storage administrators solve their replication problems, so we created a replication accelerator.
Replication acceleration using the VRX is deployed close to the application, and is focused on storage traffic only. Because the VRX is deployed close to the storage, and is targeted at helping the storage admin, we needed to find another way to redirect traffic. This led to storage-based redirection.
With storage-based redirection, all of the changes are made in the storage array or software, not the network. Traffic is sent to the VRX using a static route or a default gateway change. In effect, the VRX becomes the next hop for any replication traffic going across the WAN.
Storage-based redirection solved the deployment problem for storage administrators because it shifted the focus from the network to the storage application or array. So, problem #1 solved: how can we make this about storage and not about the network.
Solving the network problem led us down another path with regard to configuring the software and using it to monitor replication traffic.
Subtle disclaimer: I am a storage guy at a networking company. The majority of my career has been in storage, and I think about the world in storage terms and problems. When we talk about applications at Silver Peak, I immediately think of things like replication, backup, and data migration….not Exchange, or voice, or VDI. In my world, those other things just create data that needs to be stored and protected via replication and backup.
Back to the configuration and reporting.
When we started planning the next release of our Virtual Acceleration Open Architecture (VXOA), our core operating system, one of the goals was to make the VRX even easier to deploy. Additionally, we wanted to make sure that the reporting was useful for storage administrators. Now with our current release VXOA 6.2, I’m proud to say we met both of those goals. That isn’t to say that we aren’t going to do anything additional. On the contrary, I have a long list of things I would like to see in the product…but this isn’t the post for that.
With VXOA 6.2 the engineering team at Silver Peak has delivered two features that might look simple on the surface, but really deliver on the promise of building a replication accelerator that is easy to use. The new configuration wizard is only six screens, and three of them are informational. On the remaining three screens, you only enter a name, two IP addresses, and a license key. You might want to also set the date and time, and change the admin password, but you don’t have to. That’s it. Six screens, six clicks, only a few pieces of information, and you’ve configured a replication accelerator that is ready to use with nearly any replication or backup product.
The other major change is our new Data View dashboard. The Data View provides information that is key to storage administrators when it comes to monitoring their replication and backup environment across the WAN.
1. We report in storage-friendly terms, like megabytes (MB), gigabytes (GB), and terabytes (TB); not in megabits-per-second (Mbps), which is useful for networking folks. It’s easier to think about the 100 GB of data that needs to be moved every hour as a larger chunk. By the way, we left the Network View there just in case you want to know about the Mbps, loss, and latency on the WAN, or if you happen to be a network pro.
2. Inbound and outbound traffic are reported on separate graphs, which allows you to filter out the inbound traffic when you are viewing the dashboard on the sending side. You can also filter out the applications so you can see only the backup traffic, or only the replication traffic, or very specific replication traffic.
3. You can quickly determine which applications are being accelerated, and if there are any issues. The Top Applications matrix reports on the top applications and even provides throughput and deduplication information. If there are any issues, it’s a single click to move to a detailed view that provides more in-depth information and suggestions to resolve the issue.
With VXOA 6.2 Silver Peak has delivered on the promise of making replication acceleration easy and useful. The two main enhancements might not seem “fancy,” but if you’re a storage administrator trying to solve an RPO or backup problem, you don’t need fancy, you need quick, easy, and useful. VXOA 6.2 is deploys in minutes, is easy to configure, and is useful for storage traffic.
As I always say, you don’t have to believe me; you should try it in your environment, with your data, and with your WAN. Silver Peak will offer you full support during your trial. Yes, you can call our support engineers 24 hours a day, and they will help you.