Recovery-Oriented Computing (ROC)

Thomas Lindgren <>
Mon Dec 19 15:02:35 CET 2005

--- Peter-Henry Mander
<> wrote:

> Hi Joe,
> It looks very similar to your thesis [1]!
> Is there something new in the Berkeley approach?

My impression is that there has been a number of
Berkeley projects that have been transferring/updating
technology, techniques and ideas from the telco world
to the IP world, and so has Erlang. Ericsson has had a
finger in the Berkeley pie too. So maybe this overlap
is natural. (A bit strange that Berkeley hasn't heard
of Erlang, though.)

While there is a bit of commonality, there are some
system-level ideas in ROC that are not in Erlang as
such, e.g., "undo" of system states, root cause
diagnosis of faults, fault injection of various sorts,
and possibly a few others. They seem to take an
extreme approach to restarts ("micro reboots"), for
example. Some of these are arguably not
programming-language properties, of course. I guess
some of them remain to be proven to be effective too.


Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 

More information about the erlang-questions mailing list