Checkpoint/Restore track

Friday, November 4, 2016 from 2:005:00pm
Sweeney CDE
 

The MC is about Checkpoint-Restore technology in Linux and everything about it. Attendees represent the CRIU, the DMTCP, BLCR and OpenMPI developer communities.

We’ve planned talks about the C/R technology itself, about its exact implementations, about real-life usage of C/R (special guest — C/R in distributed workloads) and demoes.

Microconference Leader

Pavel Emelyanov

Proposals for this track

* What doesn't work in CRIU, ARM edition

Missing pieces for full ARM/AArch64 support and new use cases from ARM work hopefully worth sharing across all architectures
Checkpoint/Restore 09/19/2016
Christopher Covington

* Checkpoint Restore in LXD

In this talk I'll cover the ways we use checkpoint restore in LXD, as well as some challenges we face.
Checkpoint/Restore 08/04/2016
Tycho Andersen

* Checkpoint/Restart for distributed applications

Executing efficient checkpoint/restart for distributed applications
Checkpoint/Restore 09/30/2016
George Bosilca

* Lazy Process Migration

Decrease downtime during process migration using userfaultfd to restore memory pages on demand.
Checkpoint/Restore 04/29/2016
Adrian Reber, Mike Rapoport

* Some Open Challenges in Checkpointing

In this presentation, we will talk about some of the open challenges in checkpointing.
Checkpoint/Restore 09/26/2016
Kapil Arya, Gene Cooperman

* Towards better restorer

Few words about why current restorer sucks and how to make it work better.
Checkpoint/Restore 09/26/2016
Pavel Emelyanov

* What doesn't work in CRIU

Why CRIU can't dump any set of processes
Checkpoint/Restore 09/02/2016
Andrei Vagin, Adrian Reber