NetApp: Full Time Opportunities
Position Type: multiple positions
Location: Sunnyvale, CA
Overview
The WAFL file system team comprises approximately 260 Developers and QA across multiple sites in the US, Bangalore, Vancouver, and Israel. The Sunnyvale site has 80+ members of that team. This team works on critical areas like the IO path, Flash/SCM, snapshots, and has shipped critical features like Flash Pools, the All-Flash FAS (AFF), Fabric Pools, etc. The WAFL file system was created 25 years ago, but constant innovation and feature work makes it an ideal playground for software engineers that enjoy working on very difficult systems problems. Unlike other file systems that are built for a narrow use-case or workload, WAFL is engineered to handle everything from transactional workloads like databases (both in-memory & traditional) to scale-out file-based applications; the same file system can process NAS (both NFS and SMB) and SAN protocol operations since ONTAP supports multiple protocols. Over the years, dozens of features have been designed to work and interoperate within WAFL: snapshots, thin provisioning, file system cloning, compression, de-duplication, individual file cloning, compaction, software encryption, online recovery, non-disruptive file system mobility, etc. During this same time, WAFL have evolved to incorporate different media types: hard drives, flash, solid state drives, SCM, and cloud. The recent shift to a 6-month ONTAP release cadence has allowed new innovation to hit the market place in short order.
Position and Requirements
A new hire should first expect to solidify their understanding of basic file system theory by fixing some interesting bugs in core areas of WAFL before they work on a small project with guidance from a senior WAFL engineer. Then, depending on interest and abilities, they work on bigger projects with larger teams (10+ Dev and QA) or smaller 2-3 people projects. WAFL is an industrial-grade file system with a quarter million installations running business critical applications for Fortune500 as well as medium-sized corporations. There are very high expectations of data availability and resiliency even while running at million+ IOPS with sub-millisecond latencies together with dozens of features. Data unavailability is measured by our customers in millions of dollars of lost revenue. Therefore, the problems being tackled by the WAFL team are not simple, the solutions and designs are not expected to be easy, and this is certainly not for the faint of heart! A new hire should expect the work to be very difficult and the learning to be incredibly rewarding. They will find several senior engineers who are keen to share knowledge and help with this learning. Other avenues of learning include weekly chalk-talks on new and old designs, design discussions, discussion on difficult bugs, written documentation, etc. A world-class regression lab and toolset make it easy for engineers to fire off runs to measure performance improvements on industry-grade benchmarks. The ONTAP team also periodically schedules week-long hack-a-thons that allow engineers to create teams to tackle interesting problems to their liking; the only real expectation from that week is that engineers learn, have fun, and share their findings with the larger community.
How to Apply
Interested candidates can send their resume to Paul Miller, pmiller@netapp.com, director for the WAFL IO team at Sunnyvale.