SoftLayer is a different kind of cloud service provider. For one thing, it’s smaller than most, with only about 100,000 servers spread across 13 locations around the world. For another, it provides bare metal servers as a service and will move customer workloads from virtual machines to bare metal or back again.
That positions the company to offer a specialized service for users of big data. Big data customers who want to do time-sensitive analytics need not only the CPU power of a host in the cloud without virtualization’s overhead but also full access to the data transport bus and I/O ports of the server. On a multi-tenant server, users share those facilities with other users. If the analytic task commands resources frequently the other users can be shortchanged, or else the big data customer’s response times would be slowed or paused.
Wednesday at Cloud Connect, a UBM Tech event in Santa Clara, Calif., SoftLayer chief scientist Nathan Day had no comment on recent reports that IBM or EMC is interested in buying SoftLayer, but he was eager to confirm where SoftLayer is making its mark: Bare metal servers for self-provisioning big data users.
Customers using the NoSQL system MongoDB on SoftLayer’s bare metal servers prefer the service to the virtualized servers of a big, virtual machine commodity cloud supplier, Day said in an interview on the expo floor. He said customers tell him “bare metal blows away virtualized cloud servers.”
Bare metal is well suited to tasks that require frequent imports of large amounts of data, such as applying inserts and updates to the database and doing quick analyses with export of results, such as analyzing activity on a social networking or large e-commerce site. In other words, bare metal shines on big data tasks associated with lots of I/O.
“Primarily, bare metal beats virtualized servers because virtual servers use shared I/O versus dedicated I/O,” explained Harold Hannon, senior software architect at SoftLayer. Leading the effort to establish SoftLayer as a big data analytics vendor, Hannon plans to step up his blogging efforts on the company’s website advocating bare metal for analytics.
When handling large amounts of data, Hannon said, it also helps on a bare metal server if the hypervisor’s software switch doesn’t need to manage the communications inflow and outflow, which can create another slowdown point.
Through its partnership with NoSQL developer 10Gen, SoftLayer also offers MongoDB solutions. According to SoftLayer’s website, this solution provides “all the advantages the cloud has to offer but with the performance only we can provide. More RAM capacity. More Storage. More CPU speed. Bare-metal.” SoftLayer offers a version of MongoDB that can be automatically configured through an easy-to-use user interface and set up with the push of a button.
SoftLayer lets customers choose and self-provision bare-metal servers as it does virtual servers, although in a throwback to a managed application or hosted service approach, the company charges for bare metal by the month rather than by the hour. It has offered bare-metal servers since 2006.
That said, SoftLayer has not yet established a large presence among other NoSQL system providers or big data users, so it has work to do in order to become known as a differentiated service noted for big data analytics hosting capabilities. But it’s intriguing that a service provider may be carving out a niche for keeping bare metal in the cloud.
Source: Information Week