What is Live Migration compression? How does it help the migration process?
Although migration is typically regarded as a fast and reliable process, it can take several minutes or longer depending on factors such as virtual machine (VM) size, physical distance between the source and destination servers as well as bandwidth and network conditions prevailing during the migration process. For example, it can be much faster to migrate a small VM between two local data center servers across a lightly used network than to migrate a large VM between two remote servers through heavy LAN and WAN traffic.
Migration cannot control the latency between distant servers or the traffic congestion affecting the network, but it's possible to minimize the size of the VM by applying compression to the VHD or VHDX file before migration. You would then decompress the file once it arrives at the destination server.
This is the basis for Live Migration compression. By making the VM file smaller, it can be transferred between systems faster in heavy network traffic with fewer dropped or resent packets. The additional computing overhead needed to compress and decompress files is negligible on modern servers.
It's important to remember that compression works by locating and removing redundant data within a file, and replacing redundant data with small tokens that take up less space. Therefore, files with large amounts of redundant content can achieve high compression, while files with little redundant data only compress a small amount. This means different VMs will likely compress to different levels at different times, but Microsoft claims compression can generally double migration performance. IT administrators should take the time to test the effects of Live Migration compression before enabling the feature on Windows Server 2012 R2 systems in production.
This was first published in October 2013