Re:rr: lightweight recording and deterministic debugging
by roca on Tuesday November 06, 2018 @09:11AM
(#18391070)
The basic idea is the same as ReVirt, but there are a lot of different details because rr has to run as a pure user-space Linux application whereas ReVirt was baked into the hypervisor. For rr to run efficiently we have to use Linux APIs in creative ways. https://arxiv.org/abs/1705.05937 has more details.
Re:rr: lightweight recording and deterministic debugging
by MarkUndo on Wednesday November 07, 2018 @07:27AM
(#18398658)
VMware used to support record / replay of VMs in their commercial product, which was quite cool:
https://pubs.vmware.com/ws71_ace27/wwhelp/wwhimpl/js/html/ww...
Re:rr: lightweight recording and deterministic debugging
by Mic92 on Tuesday November 06, 2018 @05:12AM
(#18389719)
It is probably more mature then a research prototype as it has received more real-world testing.
Re:rr: lightweight recording and deterministic debugging
by DSingularity on Tuesday November 06, 2018 @07:35AM
(#18390412)
ReVirt logs non-deterministic inputs at the hypervisor level. RR will be recording these inputs at the Kernel level. So this makes the inability to replay the OS execution the biggest difference.
Re:rr: lightweight recording and deterministic debugging
by MarkUndo on Wednesday November 07, 2018 @07:32AM
(#18398683)
Interestingly, hypervisor-level logging of this stuff is both harder (because it has the constraints of kernel-level code) and simpler (because the non-deterministic behaviours are fewer and better-documented at the hardware level than the Linux API level!)
I think it's very likely that, overall, recording a single process is substantially more complex to implement than recording a whole VM. (With significant caveats - recording a whole VM with good performance is going to be hard and making it really useful probably is a whole load of extra code)
[1] https://www.usenix.org/legacy/events/osdi02/tech/full_papers...