> On July 1, 2013, 7:31 p.m., Benjamin Hindman wrote:
> > src/slave/flags.hpp, line 145
> > <https://reviews.apache.org/r/12084/diff/1/?file=310857#file310857line145>
> >
> >     This looks like a regression.
> 
> Benjamin Hindman wrote:
>     I didn't see you mentioned this in the description, but what was the 
> reasoning?
> 
> Vinod Kone wrote:
>     I think 'false' should be the default because I believe that catches the 
> common case, where slave dies just before writing slave info/ framework info/ 
> executor info etc. Makes sense?

I don't see how that is the common case (how often will a slave die before 
writing it's own info?), or why these are issues (if the framework/executor 
infos aren't there don't we just assume that framework or executor was never 
launched)?

Bottom line: I want to make sure that we aren't causing mysterious things to 
happen to our users forcing them to spend hours tracking down an issue that 
turns out to be an error we were silent about. When you race go karts you have 
to fill out a waiver of liability: I'd rather force our users to sign the 
'--strict=false' waiver than have them unhappy. We can always change this down 
the road with operational experience, but IIUC we don't have any yet!


- Benjamin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/12084/#review22620
-----------------------------------------------------------


On June 25, 2013, 7:16 p.m., Vinod Kone wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/12084/
> -----------------------------------------------------------
> 
> (Updated June 25, 2013, 7:16 p.m.)
> 
> 
> Review request for mesos, Benjamin Hindman and Ben Mahler.
> 
> 
> Bugs: MESOS-526
>     https://issues.apache.org/jira/browse/MESOS-526
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Also, changed the default for 'strict' from 'true' to 'false'.
> 
> 
> Diffs
> -----
> 
>   src/slave/flags.hpp 9612983b02f47d65aecf816db7048f523b2dff8f 
>   src/slave/slave.hpp 3ac893309b49e23fffebb8118d427e5d89d61a86 
>   src/slave/slave.cpp 9985f14467ca0aa7faf83e602e1d8f1c2219dfc0 
>   src/slave/state.hpp 1e3e894ee91d2b944d1d53642b5ebfcc52c02703 
>   src/slave/state.cpp 27a2731cf760f65bd90fefd38715654a63017345 
>   src/tests/slave_recovery_tests.cpp 0195e688dac09b139c107e6815e0b842116c5e5c 
> 
> Diff: https://reviews.apache.org/r/12084/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Vinod Kone
> 
>

Reply via email to