Backup, Virtualization, and the Next Big Thing: Bare Metal Cloud and Physical Servers
One of the major problems with the mantra of “virtualize everything” is the stubborn refusal of physical servers to disappear. This of course is a major irritant to virtualization companies; it’s easy to picture Pat Gelsinger at VMware screaming “die already!” at physical servers. The chart above illustrates the issue. IDC predicts that in 2016, 75% of all workloads will be virtualized compared to only 25% of workloads operating on physical servers. At the same time, IDC predicts that in 2016 hypervisor-based servers will account for only 21% of all physical servers; bare metal servers (physical servers without hypervisors) will account for 79% of all servers.
Why? The most often cited reason are tier 1 applications – mission critical applications upon which businesses are built. Quite often these applications require a tremendous amount of resources to operate and scale. Creating virtualized environments can be difficult. But tier 1 applications aren’t the only reason for so many bare metal servers; the time and expense of converting legacy applications to operate in a virtualized environment is quite often simply not worth the effort.
Even tier 1 applications that are virtualized can have a tendency to do things outside the lines of strict virtualization. A typical example is the use of physical RDM (Raw Device Mapping.) Physical RDMs are often used to decrease virtualization overhead by enabling a non-virtualized (also called physical compatibility mode) connection to underlying storage. Why should you care? Because some virtualization-only companies can’t support this type of configuration; so the tools you’re using to manage and protect your virtual environment may fail.
This is part 10 of an on-going series. Part 1: virtualization isn’t the next big thing (NBT) because it was the last big thing (LBT); part 2: data center IP traffic growth; part 3: data center IP traffic sources; part 4: cloud workloads; part 5: large data centers and administrator to server ratios; part 6: strategy overview – adapt or be crushed; part 7: automation, agility, adaptability; part 8, automation vs agility and adaptability; part 9: virtualizing everything.
