We’re off to SOSP 2017

The 26th ACM Symposium on Operating Systems Principals (SOSP 2017), a biennial event, is being held October 29-31, 2017 in Shanghai, China. VMware, a Silver sponsor, will be at the conference as well as several of our researchers who will be involved in presenting papers, leading poster sessions, and co-leading a Birds of a Feather session.

The excitement gets started a day early when the 2nd System Software for Trusted Execution (SysTEX) workshop kicks off, focusing on exploring new ideas of trustworthy systems with trusted execution environments.  Our own Nadav Amit and Michael Wei will be presenting “Hypercallbacks: A New Mechanism for Trusted, Secure Introspection.”

That evening, a birds of a feather session on “Machine Learning for All” will be held from 8-9 PM, where David Tennenhouse, VMware’s Chief Research Officer, will be co-leading the discussion with Matei Zaharia (Stanford University), Joseph Gonzalez (UC Berkeley) and Garth Gibson (Carnegie Mellon University). You do not want to miss this session. The questions being explored are: How can the systems community contribute to the democratization of AI?  Is the technology mature enough for this democratization? What systems and software are needed to support? Bring birdseed to munch on; this discussion is going to be lively.

Once SOSP kicks-off we will be presenting two papers:

Pebbles DB: Building Key-Value Stores using Fragmented Log-Structure Merge Trees

Pandian Raju, Rohan Kadekodi (University of Texas at Austin); Vijay Chidambaram (University of Texas at Austin and VMware Research); Ittai Abraham (VMware Research). The paper title almost says it all, but what it is missing is that PebblesDB outperforms RocksDB on several workloads. This is not just a cool research topic, but also has the potential for real-world applications such as Pinterest. Though you will not find it pinned to any boards, you can access the source on GitHub, https://github.com/utsaslab/pebblesdb.

Scaling a file system to many cores using an operation log

Srivatsa S. Bhat (MIT CSAIL and VMware); Rasha Eqbal (MIT CSAIL and Apple); Austin T. Clements (MIT CSAIL and Google); M. Frans Kaashoek, Nickolai Zeldovich (MIT CSAIL).  Sometimes the best ideas seem simple once you have them. This approach of decoupling the in-memory file system from the on-disk file system has shown promising results to achieve multicore scalability, durability and crash consistency in a file system. Stop by and be inspired.

But wait there’s more. The SOSP’s poster track is a chance to present ideas in their infancy to receive feedback. The poster track sessions are where you can get your crazy on and see what ideas are noodling around.  We are involved in 3 poster sessions:

Maggie Zhang, from our Beijing office, will be at our booth; you can also find Lalith Suresh, David Ott and Dan Tsafrir in attendance.  Please say hi, tell us a good joke or pick our brains about the opportunities at VMware!  We are excited to meet you.

Learn more about VMware Research: https://research.vmware.com/.  If you are interested in working for VMware Research, explore the careers site.

Other posts by

VMware Featured in ACM SIGOPS OSR Special Topics Edition

The Association for Computing Machinery (ACM) Operating Systems Review (OSR) Special Topics August 2017 edition is hot off the presses and features some of the ground-breaking research taking place within VMware. VMware is extremely honored to take part in this edition. OSR is a publication of the ACM Special Interest Group on Operating Systems (SIGOPS), whose scope […]

We Want to Meet You at SIGCOMM 2017

“All knowledge is connected to all other knowledge. The fun is in making the connections.”(Arthur C Aufderheide) Several VMware Research members are headed to the ACM SIGCOMM 2017, the ACM Special Interest Group on Data Communication (SIGCOMM 2017), which will take place August 21-25, 2017 on the UCLA campus in Los Angeles, CA. VMware is […]

Hillview: A Tool for Browsing Huge Datasets

A key part of the success of many enterprises involves analyzing and understanding data. There is a myriad of commercial and free tools available today to help enterprises do just that.  Only a few of the existing tools can handle large-scale datasets, but all of those require deep technical expertise to use. Moreover, the explosion […]