> On April 16, 2013, 1:11 a.m., Ben Mahler wrote:
> > src/tests/environment.cpp, line 71
> > <https://reviews.apache.org/r/10531/diff/1/?file=281408#file281408line71>
> >
> >     Shouldn't this be comparing against "CgroupsNoHierarchyTest"? It should 
> > probably also ensure "ROOT_" and "CGROUPS_" are present, no?
> >     
> >     That would obviate the need for you to verify the root user.
> 
> Thomas Marshall wrote:
>     Leaving off "Test" makes the logic here more flexible, since you could 
> define something like "CgroupsNoHierarchyXXXXTest" and still get the 
> filtering, but I think that a cleaner way to deal with this would be to add a 
> "CGROUPSNOHIERARCHY_" flag, since this is more consistent.
>     
>     Checking for "ROOT_" and "CGROUPS_" gets messy since its possible that 
> the flags won't appear together (i.e. some in the test name, some in the test 
> case name). So, if "ROOT_" appears in the name and "NOHIERARCHY" in the case 
> name, "NOHIERARCHY" will be checked first and getting the current hierarchies 
> will fail if the user isn't root instead of having the check for "ROOT_" 
> return false. I think that the best solution to this is to make "NOHIERARCHY" 
> also check for root and cgroups support, since there's no situation in which 
> "NOHIERARCHY" would be present that root and cgroups support wouldn't also be 
> necessary. This means that the "ROOT_" and "CGROUPS_" flags aren't necessary 
> if "NOHIERARCHY" is present, though of course there's no problem with 
> including them.

Our assumption is that all cgroups tests have "ROOT_" and "CGROUPS_ " in their 
test name (though our logic here accepts them being in test case name too). So, 
I'm ok with what you did here.

@BenM any more objections? else, i can commit this.


- Vinod


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/10531/#review19237
-----------------------------------------------------------


On April 16, 2013, 12:32 a.m., Thomas Marshall wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/10531/
> -----------------------------------------------------------
> 
> (Updated April 16, 2013, 12:32 a.m.)
> 
> 
> Review request for mesos, Benjamin Hindman and Ben Mahler.
> 
> 
> Description
> -------
> 
> I know that I'm not assigned to this issue, but its causing about 75% of the 
> test failures that we're seeing in the AMP lab's Jenkins now, so I figured I 
> would tackle it:
> 
> 
> This addresses bug MESOS-414.
>     https://issues.apache.org/jira/browse/MESOS-414
> 
> 
> Diffs
> -----
> 
>   src/tests/environment.cpp 3096e5a 
> 
> Diff: https://reviews.apache.org/r/10531/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Thomas Marshall
> 
>

Reply via email to