> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, lines 68-72
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line68>
> >
> >     Did not we describe this already?

True.  The intention was to reword the documentation in the protobufs.  But I 
think it should be ok remove this section.


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 58
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line58>
> >
> >     Why is this a Note?

I wanted it to be emphasized (that modes are all manually transitioned).  
Although, given how short this section is, I guess it isn't necessary.


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 77
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line77>
> >
> >     Here we use will, and bellow we use could. Please go over the entire 
> > doc and make sure that the use of tense is consistent. Please check other 
> > Mesos docs and see which tense do they normally use.

Looks like we should use present tense.

There are a few places where I still use future tense.  This is primarily to 
describe things like:
  The operator posts something to the endpoint.
  Master will do something in response.


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 93
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line93>
> >
> >     Is it nanonsecods since UNIX epoch? Can we just use seconds in the 
> > eaxmaple?

Yes.

I wanted the example to be correct/runnable.  So the extra 9 zeroes are 
necessary.  (I added a more reasonable timestamp, Oct 05, 2015.)


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 137
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line137>
> >
> >     what about duration? Can't we just leave it empty?

Oops, this is no longer accurate.


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 138
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line138>
> >
> >     Can't we have deactivated machines outside a maintenance schedule?

This is currently not supported.  (Because it's one of the two maintenance mode 
transitions we didn't implement.)


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 228
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line228>
> >
> >     s/even if/even when/?

I think `even if` flows better here.


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 242
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line242>
> >
> >     can we make sure that `above` is a link here?

Unfortunately, the docs on the website aren't generated with `#`-linkable 
blocks.


> On Sept. 12, 2015, 6:59 p.m., Artem Harutyunyan wrote:
> > docs/maintenance.md, line 128
> > <https://reviews.apache.org/r/37585/diff/9/?file=1061438#file1061438line128>
> >
> >     s/need to/MUST/. In general I would susggest to go over the docment and 
> > make sure we consistently use MUST,SHOULD, MAY (a la 
> > https://www.ietf.org/rfc/rfc2119.txt)

Done.

Note about "can" and "may".  I used "can" to express that something "is 
possible".  I used "may" to express something "is allowed".


- Joseph


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/37585/#review98703
-----------------------------------------------------------


On Sept. 14, 2015, 2:32 p.m., Joseph Wu wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/37585/
> -----------------------------------------------------------
> 
> (Updated Sept. 14, 2015, 2:32 p.m.)
> 
> 
> Review request for mesos, Benjamin Hindman, Ben Mahler, Artem Harutyunyan, 
> Joris Van Remoortere, and Vinod Kone.
> 
> 
> Bugs: MESOS-2083
>     https://issues.apache.org/jira/browse/MESOS-2083
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Heavily based on the design doc 
> (https://docs.google.com/document/d/16k0lVwpSGVOyxPSyXKmGC-gbNmRlisNEe4p-fAUSojk/).
> 
> Includes a diagram of the maintenance mode transitions.
> 
> 
> Diffs
> -----
> 
>   docs/images/maintenance-primitives-modes.png PRE-CREATION 
>   docs/maintenance.md PRE-CREATION 
> 
> Diff: https://reviews.apache.org/r/37585/diff/
> 
> 
> Testing
> -------
> 
> Copied to: https://gist.github.com/kaysoky/b9789c88ee204e3b49a2
> Checked for markdown correctness.
> 
> 
> File Attachments
> ----------------
> 
> Same as the image in the binary diff. (Uploaded for reviewer convenience.)
>   
> https://reviews.apache.org/media/uploaded/files/2015/09/01/7d3153ca-37f4-4948-acce-b140a3eb71a9__maintenance-primitives-modes.png
> 
> 
> Thanks,
> 
> Joseph Wu
> 
>

Reply via email to