> On Jan. 13, 2016, 11:01 p.m., Maxim Khutornenko wrote: > > src/main/java/org/apache/aurora/scheduler/AcceptedOffer.java, line 79 > > <https://reviews.apache.org/r/42126/diff/5/?file=1195816#file1195816line79> > > > > I'd expect this value populated from the command line arg supplied to > > driver setting module. Mind explaining the logic here? Why do we default to > > "*"? > > Bill Farner wrote: > Is this not necessary to align with the special handling of `*` on the > mesos side? > > Maxim Khutornenko wrote: > I guess my question was about what this default even means. I'd expect > this feature to be OFF by default by not registering with any role at all. It > should be up to cluster operator to decide what role to register with (if at > all). > > Bill Farner wrote: > In this context, i believe the logic is appropriate. Role=`*` means any > framework may use the resource. I don't think this is something you would > want to configure or disable. > https://mesos.apache.org/documentation/latest/roles/ > > However i agree with you for the role the scheduler registers as > (different part of this patch).
What is the behavior if someone decides to register with a named role, say "aurora"? Since we are sorting by RESERVED (which is DEFAULT_ROLE_NAME) first, aren't we getting the reverse behavior here? E.g. we'd get anything which is NOT "aurora" first (mostly likely only *) and then "aurora" resources? - Maxim ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/42126/#review114287 ----------------------------------------------------------- On Jan. 13, 2016, 11:50 p.m., Zhitao Li wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/42126/ > ----------------------------------------------------------- > > (Updated Jan. 13, 2016, 11:50 p.m.) > > > Review request for Aurora, Maxim Khutornenko, Dmitriy Shirchenko, and Bill > Farner. > > > Bugs: AURORA-1109 > https://issues.apache.org/jira/browse/AURORA-1109 > > > Repository: aurora > > > Description > ------- > > This review is a prototype for introducing multiple role support in Aurora. > This creates a new class OfferAllocation, which allcoates resources to > resources field in TaskInfo and ExecutorInfo from an offer. > > Current implementation prefers reserved resources over shared resources ('*' > role) if both are present > > Several caveats: > 1. This performs the allocate after scheduling decision in > TaskAssigner.maybeAssign is done, which leaves possibility of inconsistency > and late failure. > > > Diffs > ----- > > NEWS acaff9eb2ab184b0ef750f8b8a00c20131997f6b > src/main/java/org/apache/aurora/scheduler/AcceptedOffer.java PRE-CREATION > src/main/java/org/apache/aurora/scheduler/ResourceSlot.java > 7c3d681c216b78eeecebbe950186e5a79c6fe982 > src/main/java/org/apache/aurora/scheduler/Resources.java > db422a959ee7b982c2a44323de41ad75d1a40754 > > src/main/java/org/apache/aurora/scheduler/mesos/CommandLineDriverSettingsModule.java > 2255dd407cd1810c7df5baf17cfa85f79bfffeb8 > src/main/java/org/apache/aurora/scheduler/mesos/MesosTaskFactory.java > 8fdadda67478bb3110aa442b7d78493cf9c3edb4 > src/main/java/org/apache/aurora/scheduler/state/TaskAssigner.java > 7e8e456e288986eb0ce92a123b294e1e25d8ed18 > src/test/java/org/apache/aurora/scheduler/AcceptedOfferTest.java > PRE-CREATION > src/test/java/org/apache/aurora/scheduler/ResourceSlotTest.java > e4ae943303823ac4bfbe999ed22f5999484462d8 > > src/test/java/org/apache/aurora/scheduler/mesos/CommandLineDriverSettingsModuleTest.java > 33149ab415292eff04f38b61f2b1d1eac79f347a > > src/test/java/org/apache/aurora/scheduler/mesos/MesosTaskFactoryImplTest.java > a5793bffabf4e5d6195b1b99f2363d241c0cecf9 > src/test/java/org/apache/aurora/scheduler/state/TaskAssignerImplTest.java > 3cbe9acd75def14ae2e0986914ba621fb164b3e4 > > Diff: https://reviews.apache.org/r/42126/diff/ > > > Testing > ------- > > 1. Unit tested with old and new tests; > 2. vagrant integration tests: I manually separate out the vagrant box's cpu > and memory between 'aurora-test' role and '*' and verified that jobs can > still be launched (I can post the vagrant change in another follow upon > request). > > > Thanks, > > Zhitao Li > >