|
VMware enables enterprises to run multiple operating systems on server clusters in the data center, and to switch over from one operating system to another on a given server instantly. Many enterprises maintain duplicate or backup data centers and use VMware images for all mission critical servers as part of a disaster recovery strategy. System snapshots are taken using VMware. The system image can then be loaded back onto an identical or similar system in the backup data center to create an exact duplicate of that system.
VMware images are typically 500MB to Gigabytes in size, so transferring these images for many servers can be time consuming and prone to time-out errors. Even when compressed, VMware images can take up considerable bandwidth. In addition, data backup and disaster recovery can be adversely affected by latency over a WAN, particularly when using TCP. Even expensive, high bandwidth links can have difficulty supporting VMware transfers.
Silver Peak and VMware as Part of a Disaster Recovery Strategy
VMware images contain highly repetitive information and are highly compressible by Silver Peak NX Series appliances. Data centers using VMware often leverage multiple servers or server clusters running the same application, which means that there is further commonality from one image to another.
Silver Peak’s NX appliances dramatically reduce the time to backup VMware system images between disaster recovery locations.This is due to the advanced data reduction capabilities inherent to Network Memory, coupled with additional compression techniques within the Silver Peak solution.
The graph below shows a typical customer example, with Silver Peak providing a 60x improvement in VMware transfer times - from one hour to 1 minute when using a pre-compressed file.
When using uncompressed VMware images, 10x performance improvements can also be experienced on the initial image transfer.
|
