mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Anand Mazumdar" <mazumdar.an...@gmail.com>
Subject Re: Review Request 41514: Accepted a single JSON object for quota set request.
Date Mon, 21 Dec 2015 20:57:36 GMT


> On Dec. 18, 2015, 12:04 a.m., Anand Mazumdar wrote:
> > src/master/quota_handler.cpp, line 353
> > <https://reviews.apache.org/r/41514/diff/1/?file=1168675#file1168675line353>
> >
> >     hmmm .. Should we return a `BadRequest` for all other non-allowed values of
`force` other then `true` or `false` ?
> 
> Alexander Rukletsov wrote:
>     I was thinking about the [Postel's law](https://en.wikipedia.org/wiki/Robustness_principle),
but maybe you are right and we should not interpret "force:t", "force:1", "force:god-damn-yes"
as `false`, I'll fix that.

I was wondering if we should accept only `true/false` as valid values. Currently the `/scheduler`
endpoint also specifies a semantically similar field `force` in `Call::Subscribe` message
that just takes these 2 possible values. Should we strive for consistency across all the endpoints
and reject all other values ?


> On Dec. 18, 2015, 12:04 a.m., Anand Mazumdar wrote:
> > src/master/quota_handler.cpp, line 338
> > <https://reviews.apache.org/r/41514/diff/1/?file=1168675#file1168675line338>
> >
> >     I like the less jagged version more here. It's also more consistent with other
similar strings in the function. What do you think ?
> 
> Alexander Rukletsov wrote:
>     It's a known fact, that jaggedness is very subjective. I would argue, that wrapping
error message after `BadRequest` is less jagged and more readable, because more there is more
space for the message itself. Consistency is important, hence the following review: https://reviews.apache.org/r/41515/

sgtm, I dropped the issue.


- Anand


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/41514/#review111086
-----------------------------------------------------------


On Dec. 21, 2015, 1:27 p.m., Alexander Rukletsov wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/41514/
> -----------------------------------------------------------
> 
> (Updated Dec. 21, 2015, 1:27 p.m.)
> 
> 
> Review request for mesos, Anand Mazumdar, Bernd Mathiske, Joerg Schad, and Joris Van
Remoortere.
> 
> 
> Bugs: MESOS-3960
>     https://issues.apache.org/jira/browse/MESOS-3960
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> POST request to "/quota" requires a single JSON object as opposed to key-value pairs
encoded in a string.
> 
> 
> Diffs
> -----
> 
>   src/master/quota_handler.cpp 0217149a865ede751b3a03fe40b2d91b487b7b10 
>   src/tests/master_quota_tests.cpp 89130ce9b09afe7c6dd332c8b5278abe0d2674f1 
>   src/tests/role_tests.cpp 2c5f68ccaac7e9a37345e2f331d1bc35cae77736 
> 
> Diff: https://reviews.apache.org/r/41514/diff/
> 
> 
> Testing
> -------
> 
> make check on Mac OS 10.10.4
> 
> 
> Thanks,
> 
> Alexander Rukletsov
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message