Scaling track

Wednesday, September 18, 2013 from 2:004:45pm
Celestin F

The Linux Plumbers 2013 Microconference – Scaling track focuses on scalability, both upwards and downwards as well as up and down the stack. There are quite a few active projects that are working to enhance performance and scalability both in the Linux kernel and in user-space projects. In fact, one of the purposes of this forum is for developers from different projects to meet and collaborate. After all, for the user to see good performance and scalability, all relevant projects must perform and scale well.

The structure will be similar to what was followed the previous years (2011 and 2012): about 30 minutes per subject with discussion.

Microconference Leaders

Dhaval Giani, Paul McKenney

Sessions for this track

* But What About Concurrent Updates?

We need mechanisms that do for update-mostly workloads that what RCU does for read-mostly workloads. (slides)
Paul McKenney

* Current status of lock elision in Linux

Discuss the current status of lock elision in Linux (slides)
Andi Kleen

* Full dynticks status

Present state of full dynticks, current limitations and what remains to be done (slides)
Frederic Weisbecker

* Large Memory Sysfs Scaling

Scaling the userspace sysfs interface for physical memory configuration for large memory systems
Seth Jennings

* Per CPU Atomics

Implementing Per CPU atomics for userspace (slides)
Paul Turner

* Runtime NUMA Reconfiguration for Virtual Machines

Allow physical memory to change which NUMA node it belongs to at runtime to improve long running virtual machine performance. (slides)
Cody Schafer

* Scalability Issues in Linux Kernel

Scalability Issues in Linux Kernel (slides)
Tim Chen, Dave Hansen

* User-level threads....... with threads.

New Scheduler support allowing cooperative scheduling on top of a a 1:1 threading model. (slides)
Paul Turner

* Userspace RCU library: new APIs and data structures

Discussion of userspace RCU new APIs and use-cases (slides)
Mathieu Desnoyers

* vmsplice with transparent huge pages

vmsplice() could move (rather can copy) pages between processes, but performance would be greatly improved if this supported THP.
Robert Jennings

Proposals for this track

* Enhancing syncronization mechanisms: mutexes and rwsems

An overview of recent locking enhancements for mutexes and rwsems.
Scaling 07/31/2013
Davidlohr Bueso

* Maximizing VM writeout performance

Teach the VM subsystem to perform large I/O writeouts to avoid performance degradation in some situations.
Scaling 07/15/2013
Robert Jennings

* Scaling Monitoring and Discovery above the 100K systems range

We give an overview of algorithms for reliable exception monitoring and control that extends beyond the 100K system range while minimizing network traffic.
Scaling 08/02/2013
Alan Robertson