> On 一月 28, 2016, 1:29 p.m., Guangya Liu wrote: > > docs/quota.md, line 279 > > <https://reviews.apache.org/r/42910/diff/1/?file=1224950#file1224950line279> > > > > I think that the current logic is "A quota'ed role is only allocated > > the reserved resources beyond its quota"? Because I see that the current > > logic is both non reserved and non revocable which should count in non > > reserved resources are not allocated. > > > > > > https://github.com/apache/mesos/blob/master/src/master/allocator/mesos/hierarchical.cpp#L1346-L1349 > > Alexander Rukletsov wrote: > I'm not sure I got your comment : ). Do you think the note is misleading? > The only fact that is not entirely correct is that even if a quota'ed > framework opts for revocable resources, currently it won't receive any > additional revocable resource beyond quota. However since we plan to change > it in the next release I decided to leave the sentence there so we won't > forget it : ).
It is just nit, from what I specified in the link https://github.com/apache/mesos/blob/master/src/master/allocator/mesos/hierarchical.cpp#L1346-L1349 , not only the non-revocable resources (Already count in non reserved resources), but also the non reserved resources are not allocated, but here only mentionoing non-revocable. A more accurate and clear sentences should be either "A quota'ed role may not be allocated any non-revocable and non-reserved resources" or "A quota'ed role is only allocated the reserved resources beyond its quota". If I was the reader of this document, then with "A quota'ed role may not be allocated any non-revocable resources", I will think what about non-reserved resources, does the non-reserved resources will be allocated? I cannot get any answer from the current document if the non reserved resources are allocated beyond quota. - Guangya ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/42910/#review116769 ----------------------------------------------------------- On 一月 28, 2016, 1:07 p.m., Alexander Rukletsov wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/42910/ > ----------------------------------------------------------- > > (Updated 一月 28, 2016, 1:07 p.m.) > > > Review request for mesos, Joris Van Remoortere and Michael Park. > > > Repository: mesos > > > Description > ------- > > See summary. > > > Diffs > ----- > > docs/quota.md 2762557ad1cec0655a27c4dda4016e07c7f63cd9 > > Diff: https://reviews.apache.org/r/42910/diff/ > > > Testing > ------- > > None. > > > Thanks, > > Alexander Rukletsov > >
