> On May 26, 2015, 8:57 p.m., Vinod Kone wrote:
> > src/common/resources.cpp, line 1160
> > <https://reviews.apache.org/r/34676/diff/1/?file=972037#file972037line1160>
> >
> > Lets use something other than "*", since we already abuse it to
> > represent the default role and unreserved resources.
> >
> > How about "R"?
>
> Brian Wickman wrote:
> I like "*" because it connotes a tainted number a la record books:
> http://en.wikipedia.org/wiki/Asterisk#Competitive_sports_and_games
>
> I'm fine with something other than * though I'm not sure 'R' is the best
> either.
>
> Vinod Kone wrote:
> I see. Since "*" is currently at the end of the resource representation,
> maybe it's not that confusing.
>
> My only concern is that when we start having extra bits of information in
> RevocableInfo, we likely want to pull that to the left of ":" (e.g, "disk
> (role, reservationinfo) [diskinfo] {revocableinfo} : 10"), at which point a
> "*" becomes confusing).
>
> Thoughts?
>
> Brian Wickman wrote:
> in the absence of any information in RevocableInfo at the moment, would
> you just have "cpu(hadoop){}:10" to represent oversubscribed cpu?
I would vote for "cpu(hadoop){REV}:10" for now. Later on when we have more info
in RevocableInfo we can update the log string.
- Vinod
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/34676/#review85217
-----------------------------------------------------------
On May 26, 2015, 11:19 p.m., Brian Wickman wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/34676/
> -----------------------------------------------------------
>
> (Updated May 26, 2015, 11:19 p.m.)
>
>
> Review request for mesos and Vinod Kone.
>
>
> Bugs: MESOS-2693
> https://issues.apache.org/jira/browse/MESOS-2693
>
>
> Repository: mesos
>
>
> Description
> -------
>
> Add printing of extended attributes to Resource objects. I have no idea what
> format is appropriate. I made something that stands some chance of being
> parseable in the future.
>
>
> Diffs
> -----
>
> src/common/resources.cpp d93f38ecdc31605b6a84c8d103ac90e8f0632b01
> src/tests/resources_tests.cpp 459a68bb8e30852becc7d8825a24f984cb027b08
>
> Diff: https://reviews.apache.org/r/34676/diff/
>
>
> Testing
> -------
>
> make check
>
>
> Thanks,
>
> Brian Wickman
>
>