Technology has its share of unsung heroes. Wearing hoodies instead of capes, these folks work quietly in the background, solving seemingly small problems that can be huge hurdles to progress.
One such group at VMware recently found answers to a couple of important problems involving time synchronization. First off, Michael Wei (along with Ali Najafi, who is now at Meta) demonstrated how to use everyday computer clocks for precise time synchronization (we’re talking microseconds and nanoseconds here). This precision is necessary for large distributed databases and multi-cloud applications to function optimally. By dispelling the notion that local computer clocks are unreliable (and being able to predict the degree of drift), their work could save companies millions of dollars they’d otherwise be spending for app developers to build more complicated algorithms for synchronization and consistency. The findings were important enough to win them “best paper” at a recent USENIX Symposium on Networked Systems Design and Implementation (NSDI).
Another group — which included Radhika Niranjan Mysore, Lalith Suresh, Gerd Zellweger, Bo Gan, Timothy Merrifield, and Sujata Banerjee, together with Timothy Roscoe and Roni Haecki from ETH Zurich — recently solved another time problem that has vexed multi-cloud software developers: pinpointing the exact causes of network latency within host network stacks. They also presented their findings at NSDI. You can read more detail about their research in this blog.
These groups are two among many at VMware whose research focuses on foundational enablement of subsequent innovation, such as new apps or services, increased performance, or simplified app development. So, here’s to all the humble heroes in our midst!