Re: Review Request 51028: Fix tests with rapidly triggered allocations.

--- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/51028/#review149316 --- Jacob, can you help rebase? - Guangya Liu On 九月 8, 2016, 10:12

Re: Review Request 51621: WIP: Made recovered resource allocated as soon as possible.

SlaveLost" --gtest_repeat=20 GLOG_v=1 ./bin/mesos-tests.sh --gtest_filter="SlaveTest.ExecutorShutdownGracePeriod" --gtest_repeat=20 ``` Thanks, Guangya Liu

Review Request 51946: Updated test for BadACLNoPrincipal and BadACLDropCreateAndDestroy.

="*PersistentVolumeTest.BadACL*/*" --gtest_repeat=20 ``` Thanks, Guangya Liu

Review Request 51942: Removed scalars() API.

add48c7baf8fabf5cc443d60a6c96d2902fc67de src/common/resources.cpp aa4d6b5af9c24e90cefaa4f4c1ad593966eceb7c src/v1/resources.cpp 4b66acbba92bc6316fa77ae4dcabce9fab65f7b5 Diff: https://reviews.apache.org/r/51942/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 51885: Added a basic default executor implementation.

> On 九月 14, 2016, 1:08 p.m., Guangya Liu wrote: > > src/launcher/default_executor.cpp, line 203 > > <https://reviews.apache.org/r/51885/diff/1/?file=1498346#file1498346line203> > > > > s/multiple task groups /multiple tasks in task group > > Vinod Kon

Re: Review Request 51877: Updated comments for += and -= `Resource_` object.

) - include/mesos/resources.hpp 7ba422d57d4e058c682df9aa780557782423 include/mesos/v1/resources.hpp add48c7baf8fabf5cc443d60a6c96d2902fc67de Diff: https://reviews.apache.org/r/51877/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 51877: Updated comments for += and -= `Resource_` object.

ed to Resource_` in comments. - Guangya --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/51877/#review148940 ----------- On

Re: Review Request 51683: Avoided resource validation when flatten resources.

nvironment tear-down [==] 1 test from 1 test case ran. (24 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Re: Review Request 51885: Added a basic default executor implementation.

> On 九月 14, 2016, 1:08 p.m., Guangya Liu wrote: > > How about rename this file as `default_pod_executor.cpp`? > > Vinod Kone wrote: > we are calling it "pod executor" because this executor will be able to > launch pod and non-pods in the future. command

Re: Review Request 51886: Added tests for default executor.

> On 九月 14, 2016, 1:41 p.m., Guangya Liu wrote: > > src/tests/default_executor_tests.cpp, lines 119-123 > > <https://reviews.apache.org/r/51886/diff/1/?file=1498340#file1498340line119> > > > > How about adding another task to the task group? ``` v1::TaskInf

Re: Review Request 51886: Added tests for default executor.

tps://reviews.apache.org/r/51886/#comment216405> How about adding another task to the task group? - Guangya Liu On 九月 14, 2016, 9:34 a.m., Anand Mazumdar wrote: > > --- > This is an automatically generated e-mail. To reply,

Re: Review Request 51885: Added a basic default executor implementation.

ent216404> How about rename this as `DefaultPodExecutor`? src/launcher/default_executor.cpp (line 203) <https://reviews.apache.org/r/51885/#comment216403> s/multiple task groups /multiple tasks in task group - Guangya Liu On 九月 14, 2016, 9:34 a.m., An

Re: Review Request 51856: Updated agent to correctly populate `CommandInfo` for default executor.

1022) <https://reviews.apache.org/r/51856/#comment216393> s/string/std::string - Guangya Liu On 九月 13, 2016, 9:20 p.m., Anand Mazumdar wrote: > > --- > This is an automatically generated e-mail. To reply,

Re: Review Request 49617: Add benchmark for failover of many frameworks.

> On 八月 24, 2016, 9:12 a.m., Guangya Liu wrote: > > src/tests/hierarchical_allocator_tests.cpp, lines 4091-4099 > > <https://reviews.apache.org/r/49617/diff/7/?file=1476268#file1476268line4091> > > > > Can we make sure all of the `allocations` are available

Re: Review Request 51683: Avoided resource validation when flatten resources.

3 ms total) [--] Global test environment tear-down [==] 1 test from 1 test case ran. (24 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Re: Review Request 51621: WIP: Made recovered resource allocated as soon as possible.

os-tests.sh --gtest_filter="MasterAllocatorTest/1.RebalancedForUpdatedWeights" --gtest_repeat=30 GLOG_v=1 ./bin/mesos-tests.sh --gtest_filter="HierarchicalAllocatorTest.UpdateWeight" --verbose --gtest_repeat=20 GLOG_v=1 ./bin/mesos-tests.sh --gtest_filter="MasterAllocatorTest/*.SlaveLost" --gtest_repeat=20 ``` Thanks, Guangya Liu

reviews@mesos.apache.org

> On 九月 14, 2016, 12:08 a.m., Guangya Liu wrote: > > include/mesos/resources.hpp, lines 511-512 > > <https://reviews.apache.org/r/51850/diff/1/?file=1497888#file1497888line511> > > > > I think that we should also mention for how to convert a `Resource` &g

Review Request 51877: Updated comments for += and -= `Resource_` object.

/resources.hpp 7ba422d57d4e058c682df9aa780557782423 include/mesos/v1/resources.hpp add48c7baf8fabf5cc443d60a6c96d2902fc67de Diff: https://reviews.apache.org/r/51877/diff/ Testing --- make make check Thanks, Guangya Liu

reviews@mesos.apache.org

tps://reviews.apache.org/r/51850/#comment216321> I think that we should also mention for how to convert a `Resource` object to `Resource_` object in the comments of this API, otherwise, people may not know how to add a `Resource` object. - Guangya Liu On 九月 13, 2016, 5:04 p.m., Jiang Yan Xu

Re: Review Request 51771: Updated master validation for DOCKER ContainerInfo in TaskGroup.

src/tests/master_validation_tests.cpp (lines 1917 - 1918) <https://reviews.apache.org/r/51771/#comment216193> new line here - Guangya Liu On 九月 13, 2016, 10:31 a.m., Abhishek Dasgupta wrote: > > --- > This is an auto

Re: Review Request 51027: Track allocation candidates to bound allocator.

- On 九月 3, 2016, 6:05 a.m., Jacob Janco wrote: > > --- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/51027/ >

Re: Review Request 51816: Updated the upgrades.md and CHANGELOG for MESOS-6150.

`deprecated` may cause confusion: `deprecated` does not always mean the API is removed but sometimes may mean this API is in a deprecating cycle. docs/upgrades.md (line 203) <https://reviews.apache.org/r/51816/#comment216176> Ditto as above, `deprecated` may not fit well in this sce

Re: Review Request 51780: Changed isolator recover interface using 'ContainerRecoverInfo'.

tps://reviews.apache.org/r/51780/#comment216177> Just a question here, does this interface need a deprecation cycle? This may cause some customized isolators does not work. - Guangya Liu On 九月 12, 2016, 6:41 p.m., Gilbert Song

Re: Review Request 51780: Changed isolator recover interface using 'ContainerRecoverInfo'.

https://reviews.apache.org/r/49234/ for a full clean up. - Guangya Liu On 九月 12, 2016, 6:41 p.m., Gilbert Song wrote: > > --- > This is an automatically generated e-mail. To reply, visit: > https://reviews.

Re: Review Request 51771: Updated master validation for DOCKER ContainerInfo in TaskGroup.

> On 九月 10, 2016, 4 a.m., Guangya Liu wrote: > > src/tests/master_validation_tests.cpp, line 1722 > > <https://reviews.apache.org/r/51771/diff/1/?file=1495218#file1495218line1722> > > > > s/ContainerInfo Type/`ContainerInfo` > > s/DOCKER/`DOCKER` &

Re: Review Request 51778: Fixed OSX build by adding lambda function stout header.

/ - Guangya Liu On 九月 10, 2016, 10:13 p.m., Gilbert Song wrote: > > --- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache

Re: Review Request 51683: Avoided resource validation when flatten resources.

nvironment tear-down [==] 1 test from 1 test case ran. (24 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Re: Review Request 51683: Avoided resource validation when flatten resources.

Thanks Jiang Yan, I made an update for all of the non-testing code, hope it is OK. - Guangya --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/51683/#review148302 -------

Re: Review Request 51771: Updated master validation for DOCKER ContainerInfo in TaskGroup.

idation_tests.cpp (line 1894) <https://reviews.apache.org/r/51771/#comment215852> I think that we can kill `Type` here src/tests/master_validation_tests.cpp (line 1928) <https://reviews.apache.org/r/51771/#comment215853> s/ContainerInfo Type/`ContainerInfo` s/DOCKER

Re: Review Request 49617: Add benchmark for failover of many frameworks.

> On 九月 9, 2016, 7:06 a.m., Guangya Liu wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 4035 > > <https://reviews.apache.org/r/49617/diff/7/?file=1476268#file1476268line4035> > > > > You can call ` SlaveInfo createSlaveInfo(const Resources& re

Re: Review Request 49617: Add benchmark for failover of many frameworks.

> > When allocations triggered by `deactivateFramework` are running, the > > `foreach` on `allocations` doesn't work. > > > > `recoverResources` doesn't trigger allocations (yet) so we don't need > > to settle after the loop. This may change but let'

Re: Review Request 51028: Fix tests with rapidly triggered allocations.

ke the code not cosistent, what is your concern if we continue using `offerCallbacks` here? - Guangya Liu On 九月 8, 2016, 10:12 p.m., Jacob Janco wrote: > > --- > This is an automatically generated e-mail. To reply,

Re: Review Request 51028: Fix tests with rapidly triggered allocations.

t the end src/tests/hierarchical_allocator_tests.cpp (line 3603) <https://reviews.apache.org/r/51028/#comment215793> 1) The `allocationRuns.load()` actually include the number of allocations caused by adding agent, we should remove this number here. 2) kill the period at the end -

Re: Review Request 49617: Add benchmark for failover of many frameworks.

> On 九月 8, 2016, 5:24 p.m., Jiang Yan Xu wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 4065 > > > > > > Move settle to above the loop. > > > > When allocations triggered by `deactivateFramewor

Re: Review Request 49617: Add benchmark for failover of many frameworks.

tps://reviews.apache.org/r/49617/#comment215790> You can call ` SlaveInfo createSlaveInfo(const Resources& resources)` so as to avoid calling `Resources::parse()` each time. - Guangya Liu On 八月 17, 2016, 2:26 a.m., Jacob J

Re: Review Request 49617: Add benchmark for failover of many frameworks.

> On 八月 24, 2016, 9:12 a.m., Guangya Liu wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 4030 > > <https://reviews.apache.org/r/49617/diff/7/?file=1476268#file1476268line4030> > > > > How about adding `ports` resources to the allocation here. >

Review Request 51753: Fixed build error for mesos agent provisioner.

Description --- Fixed build error for mesos agent provisioner. Diffs - src/slave/containerizer/mesos/provisioner/paths.cpp 68bc1b4e8252e2b1653dbab5ad58ac0b7a770bac Diff: https://reviews.apache.org/r/51753/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 51028: Fix tests with rapidly triggered allocations.

> On 九月 6, 2016, 9:59 a.m., Guangya Liu wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 2594 > > <https://reviews.apache.org/r/51028/diff/2/?file=1476333#file1476333line2594> > > > > Seems we should also `Clock::settle();` here, otherwise, this t

Re: Review Request 51027: Track allocation candidates to bound allocator.

> On 九月 3, 2016, 9:21 a.m., Guangya Liu wrote: > > src/master/allocator/mesos/hierarchical.cpp, lines 484-485 > > <https://reviews.apache.org/r/51027/diff/4/?file=1490734#file1490734line484> > > > > I did some test with benchmark tes

Re: Review Request 51683: Avoided resource validation when flatten resources.

--] Global test environment tear-down [==] 1 test from 1 test case ran. (24 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Re: Review Request 51477: Implemented `RunTaskGroupMessage` handler on the agent.

> On 九月 7, 2016, 10:17 a.m., Guangya Liu wrote: > > src/slave/slave.hpp, line 1027 > > <https://reviews.apache.org/r/51477/diff/3/?file=1492260#file1492260line1027> > > > > How about const? > > Anand Mazumdar wrote: > Can you elaborate? Non memb

Re: Review Request 51477: Implemented `RunTaskGroupMessage` handler on the agent.

``` src/slave/slave.cpp (lines 2079 - 2080) <https://reviews.apache.org/r/51477/#comment215576> new line here src/slave/slave.cpp (line 2085) <https://reviews.apache.org/r/51477/#comment215577> How about also adding `[]` here? ``` out << "tasks ["

Re: Review Request 51477: Implemented `RunTaskGroupMessage` handler on the agent.

> On 九月 7, 2016, 10:17 a.m., Guangya Liu wrote: > > src/slave/slave.hpp, line 1027 > > <https://reviews.apache.org/r/51477/diff/3/?file=1492260#file1492260line1027> > > > > How about const? > > Anand Mazumdar wrote: > Can you elaborate? Non memb

Re: Review Request 50695: Renamed agent used resources to `allocation` for benchmark test.

ck, thanks Jiang Yan. - Guangya --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/50695/#review148016 --- On 九月 8, 2

Re: Review Request 50695: Renamed agent used resources to `allocation` for benchmark test.

ms total) ``` Thanks, Guangya Liu

Re: Review Request 51480: Added tests for launching task groups on the agent.

g/r/51480/#comment215369> s/taskgroup/task group src/tests/slave_tests.cpp (line 4608) <https://reviews.apache.org/r/51480/#comment215305> 4 spaces - Guangya Liu On 九月 6, 2016, 9:36 p.m., Anand Mazumdar wrote: > > --- &g

Re: Review Request 51478: Modified tests to account for `_runTask()` rename.

4 spaces - Guangya Liu On 九月 6, 2016, 9:27 p.m., Anand Mazumdar wrote: > > --- > This is an automatically generated e-mail. To reply, visit: > https://re

Re: Review Request 51477: Implemented `RunTaskGroupMessage` handler on the agent.

://reviews.apache.org/r/51478/ , this patch will cause build failed. That's to say this patch is not atomic and I think that mesos always want an atomic patch. But it also ok if those two patches can be merged together. - Guangya Liu On 九月 6, 2016, 9:25 p.m., Anand Mazumdar

Re: Review Request 51477: Implemented `RunTaskGroupMessage` handler on the agent.

oduce a new `TASKGROUP_KILLED` for this? Ditto for other places. src/slave/slave.cpp (lines 3099 - 3114) <https://reviews.apache.org/r/51477/#comment215346> Can you please show more detail and update the comments here for which case will cause the `executor->queuedTasks` and `exe

Re: Review Request 51621: WIP: Made recovered resource allocated as soon as possible.

; --gtest_repeat=20 GLOG_v=1 ./bin/mesos-tests.sh --gtest_filter="ReservationEndpointsTest.ReserveAvailableAndOfferedResources" --gtest_repeat=20 GLOG_v=1 ./bin/mesos-tests.sh --gtest_filter="ReservationTest.ACLMultipleOperations" --verbose --gtest_repeat=20 ``` Thanks, Guangya Liu

Re: Review Request 50836: Made add/subtract resource object as private method.

> > > The `friend` keyword wouldn't allow you to specify `static` but here > > you can. > > > > Then in the cpp files you don't have to remove the `static` keyword on > > the definition. > > Guangya Liu wrote: > Thanks Jiang Yan

Review Request 51683: Avoided resource validation when flatten resources.

/resources_tests.cpp f627862870279e5a778aa83e1ddcaa88c27ba1cb src/tests/role_tests.cpp 162c9414a70723a212cfd39ac5c0b8325c3f3b5d src/v1/resources.cpp 4ba5bf87c9b9fef67eeb447a3f3b520f1e81ad77 Diff: https://reviews.apache.org/r/51683/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 50836: Made add/subtract resource object as private method.

> > > The `friend` keyword wouldn't allow you to specify `static` but here > > you can. > > > > Then in the cpp files you don't have to remove the `static` keyword on > > the definition. > > Guangya Liu wrote: > Thanks Jiang Yan

Re: Review Request 50695: Renamed agent used resources to `allocation` for benchmark test.

(6747 ms total) ``` Thanks, Guangya Liu

Re: Review Request 51028: Fix tests with rapidly triggered allocations.

tps://reviews.apache.org/r/51028/#comment215082> Seems we should also `Clock::settle();` here, otherwise, this test may be failed as r/51027 - Guangya Liu On 八月 23, 2016, 8:47 a.m., Jacob Janco wrote: > > --- > This is a

Re: Review Request 51028: Fix tests with rapidly triggered allocations.

> On 九月 5, 2016, 11:23 p.m., Jiang Yan Xu wrote: > > src/tests/hierarchical_allocator_tests.cpp, lines 2718-2720 > > > > > > How about: > > > > ``` > > // Wait for the allocation triggered from `addSl

Re: Review Request 51621: WIP: Made recovered resource allocated as soon as possible.

: https://reviews.apache.org/r/51621/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 51421: Added provisioner appc unit test for recovering nested container.

ng/const string src/tests/containerizer/provisioner_appc_tests.cpp (lines 541 - 546) <https://reviews.apache.org/r/51421/#comment214976> What about moving this right before #551? Ditto for #553 to #556 - Guangya Liu On 八月 29, 2

Re: Review Request 51420: Added provisioner appc unit test for provisioning nested container.

383 - 386) <https://reviews.apache.org/r/51420/#comment214963> What about moving this to the end of the test just before when it was used before #404? - Guangya Liu On 八月 29, 2016, 9:20 p.m., Gilbert Song wrote: > > --- > T

Re: Review Request 51402: Added nested container check in provisioner destroy.

e sub-containers were already destroyed, so in the followig check, it shoud use check-fail instead? - Guangya Liu On 八月 29, 2016, 9:20 p.m., Gilbert Song wrote: > > --- > This is an automatically generated e-mail

Re: Review Request 51393: Added unit test for provisioner recursive listContainers().

less jagged? ``` // TODO(gilbert): Refactor duplicate code // to the unit test class. ``` - Guangya Liu On 八月 29, 2016, 9:20 p.m., Gilbert Song wrote: > > --- > This is an automatically generated e-mail. To reply, visit: > http

Review Request 51621: Made recovered resource allocated as soon as possible.

if there are no request in the message queue. Diffs - src/master/allocator/mesos/hierarchical.cpp 9e5db2196c6a541dc1208ba8b9f13ef9a518bcc4 Diff: https://reviews.apache.org/r/51621/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 51027: Track allocation candidates to bound allocator.

> On 九月 3, 2016, 9:21 a.m., Guangya Liu wrote: > > src/master/allocator/mesos/hierarchical.cpp, lines 484-485 > > <https://reviews.apache.org/r/51027/diff/4/?file=1490734#file1490734line484> > > > > I did some test with benchmark tes

Re: Review Request 51027: Track allocation candidates to bound allocator.

kCount/HierarchicalAllocator_BENCHMARK_Test.AddAndUpdateSlave/0, where GetParam() = (1000, 1) 1 FAILED TEST ``` The reason is that we cannot make sure all of the `_allocate()` finished after `addSlave` finished. Shall we do a `while` loop in the benchmark to wait till all allocations are got? - Guangya Liu O

Re: Review Request 50868: Using `OfferedResources` for benchmark test of `SuppressOffers`.

; > need to change the tests to use offers directly but right now I think 1) > > "allocation" is most correct, 2) it's not a big deal if people treat them > > as synonymous terms. > > > > That said, I think reusing the "Allocation" defined a

Re: Review Request 49617: Add benchmark for failover of many frameworks.

> On 八月 24, 2016, 9:12 a.m., Guangya Liu wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 4030 > > <https://reviews.apache.org/r/49617/diff/7/?file=1476268#file1476268line4030> > > > > How about adding `ports` resources to the allocation here. >

Re: Review Request 51027: Track allocation candidates to bound allocator.

uot;Performed deallocation for " << allocationCandidates.size() << " agents in " << stopwatch.elapsed(); ``` - Guangya Liu On 九月 3, 2016, 6:05 a.m., Jacob Janco wrote: > > --

Re: Review Request 50693: Added a `createSlaveInfo()` overload that takes a `Resources`.

14140 ms total) ``` Thanks, Guangya Liu

Re: Review Request 51359: Added unit test for provisioner helper findContainerDir.

<https://reviews.apache.org/r/51359/#comment214878> How about ``` EXPECT_SOME_EQ( containerDir1, findContainerDir(provisionerDir, child1).get()); ``` and ditto for others - Guangya Liu On 八月 29, 2016, 9:20 p.m., Gilber

Re: Review Request 51359: Added unit test for provisioner helper findContainerDir.

) <https://reviews.apache.org/r/51359/#comment214874> move this right before #52 where it was used first? - Guangya Liu On 八月 29, 2016, 9:20 p.m., Gilbert Song wrote: > > --- > This is an automatically generated e-mail.

Re: Review Request 51358: Implemented recursive helper method findContainerDir for provisioner.

) <https://reviews.apache.org/r/51358/#comment214871> kill this src/slave/containerizer/mesos/provisioner/paths.cpp (lines 132 - 133) <https://reviews.apache.org/r/51358/#comment214873> How about putting the `container directory path` into the error message? - Guangya Li

Re: Review Request 51553: Changed the way `HAP::updateAllocation()` calls `Resources.apply()`.

h time. - Guangya Liu On 九月 1, 2016, 4:35 p.m., Jiang Yan Xu wrote: > > --- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/51553/ > ---

Re: Review Request 51027: Track allocation candidates to bound allocator.

related to `shared resources` are now merged in `hierarchical.cpp`? Thanks. - Guangya Liu On 八月 30, 2016, 4:53 p.m., Jacob Janco wrote: > > --- > This is an automatically generated e-mail. To reply, visit: > https://reviews.

Re: Review Request 50693: Added another function `createSlaveInfo` for allocator benchmark test.

/HierarchicalAllocator_BENCHMARK_Test (14140 ms total) ``` Thanks, Guangya Liu

Re: Review Request 50693: Added another function `createSlaveInfo` for allocator benchmark test.

/HierarchicalAllocator_BENCHMARK_Test (14140 ms total) ``` Thanks, Guangya Liu

Re: Review Request 50693: Added another function `createSlaveInfo` for allocator benchmark test.

e.org/r/49571/ that does this because we need to > > construct more complex resources. Will it be OK to just have this method in > > https://reviews.apache.org/r/49571/? > > Guangya Liu wrote: > Thanks Jiang Yan, what about make the patch of > https://reviews.ap

Re: Review Request 50693: Added another function `createSlaveInfo` for allocator benchmark test.

ally generated e-mail. To reply, visit: https://reviews.apache.org/r/50693/#review147603 ----------- On 八月 26, 2016, 7:56 a.m., Guangya Liu wrote: > > --- >

Re: Review Request 50836: Made add/subtract resource object as private method.

> > > The `friend` keyword wouldn't allow you to specify `static` but here > > you can. > > > > Then in the cpp files you don't have to remove the `static` keyword on > > the definition. > > Guangya Liu wrote: > Thanks Jiang Yan

Re: Review Request 51528: Added 'HierarchicalAllocatorTest.ResourceMetricsUninitialized' test.

> On 八月 30, 2016, 6:41 p.m., Guangya Liu wrote: > > @Jiang Yan, what about switch the order of this and > > https://reviews.apache.org/r/51529/ so as to make sure the build will not > > be failed after this merged? Sorry, I mean the unit test will failed if keeping su

Re: Review Request 51528: Added 'HierarchicalAllocatorTest.ResourceMetricsUninitialized' test.

://reviews.apache.org/r/51529/ so as to make sure the build will not be failed after this merged? - Guangya Liu On 八月 30, 2016, 5:07 p.m., Jiang Yan Xu wrote: > > --- > This is an automatically generated e-mail. To reply, visit

Re: Review Request 51027: Track allocation candidates to bound allocator.

is processed, the set of candidates is cleared. ``` - Guangya Liu On 八月 30, 2016, 4:53 p.m., Jacob Janco wrote: > > --- > This is an automatically generated e-mail. To reply, visit: >

Re: Review Request 50868: Using `OfferedResources` for benchmark test of `SuppressOffers`.

; > need to change the tests to use offers directly but right now I think 1) > > "allocation" is most correct, 2) it's not a big deal if people treat them > > as synonymous terms. > > > > That said, I think reusing the "Allocation" defined a

Review Request 51481: Updated TaskGroupValidationTest.TaskUsesNetworkInfo use two tasks.

[ OK ] TaskGroupValidationTest.TaskUsesNetworkInfo (143 ms) [--] 1 test from TaskGroupValidationTest (143 ms total) [--] Global test environment tear-down [==] 1 test from 1 test case ran. (153 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Review Request 51472: Added test case TaskGroupValidationTest.TaskUsesExecutor.

esExecutor [ OK ] TaskGroupValidationTest.TaskUsesExecutor (91 ms) [--] 1 test from TaskGroupValidationTest (91 ms total) [--] Global test environment tear-down [==] 1 test from 1 test case ran. (101 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Review Request 51455: Fixed typo for the comments of UnauthorizedTaskGroup.

--- Fixed typo for the comments of UnauthorizedTaskGroup. Diffs - src/tests/master_authorization_tests.cpp d0fe4bed24077d68cc31a3ea031bb9b04a14165d Diff: https://reviews.apache.org/r/51455/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 51425: Added missed headers.

/docker.hpp (line 22) <https://reviews.apache.org/r/51425/#comment213774> kill this, no one is using this. src/slave/containerizer/docker.hpp (line 24) <https://reviews.apache.org/r/51425/#comment213775> kill this, you can add this back when your patch does need this. - Guangya

Re: Review Request 51451: Added test case MasterAuthorizationTest.KillPendingTaskInTaskGroup.

ent tear-down [==] 1 test from 1 test case ran. (100 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Re: Review Request 50868: Using `OfferedResources` for benchmark test of `SuppressOffers`.

, they are focusing on `allocations` for each framework, so it is using `Allocation`. What do you think? - Guangya --- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/50868/#review146878 --

Review Request 51451: Added test case MasterAuthorizationTest.KillPendingTaskInTaskGroup.

` Thanks, Guangya Liu

Review Request 51447: Killed a line in master/validation.cpp.

--- Killed a line in master/validation.cpp. Diffs - src/master/validation.cpp 3d2965e4cfaae075f988fd43386bd00c7d807b2e Diff: https://reviews.apache.org/r/51447/diff/ Testing --- make make check Thanks, Guangya Liu

Re: Review Request 51437: Added validation test for task group task using NetworkInfos.

tps://reviews.apache.org/r/51437/#comment213668> Here also needs an update if you quota `NetworkInfos`. - Guangya Liu On 八月 25, 2016, 7:34 p.m., Vinod Kone wrote: > > --- > This is an automatically generated e-mail. To reply,

Re: Review Request 51437: Added validation test for task group task using NetworkInfos.

bout define a variable first and use this `resource` in both executor and task (#1784)? - Guangya Liu On 八月 25, 2016, 7:34 p.m., Vinod Kone wrote: > > --- > This is an automatically generated e-mail. To reply, vi

Review Request 51433: Added test case MasterAuthorizationTest, UnauthorizedTaskGroup.

Test [ RUN ] MasterAuthorizationTest.UnauthorizedTaskGroup [ OK ] MasterAuthorizationTest.UnauthorizedTaskGroup (85 ms) [--] 1 test from MasterAuthorizationTest (86 ms total) [--] Global test environment tear-down [==] 1 test from 1 test case ran. (105 ms total) [ PASSED ] 1 test. ``` Thanks, Guangya Liu

Review Request 51432: Added test case for TaskGroupValidationTest.ExecutorWithoutFrameworkId.

] 1 test. ``` Thanks, Guangya Liu

Re: Review Request 50841: Added GPU scheduling logic to docker containerizer process.

automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/50841/#review146726 --- On 八月 22, 2016, 10:11 a.m., Yubo Li wrote: > > --- > This is an automatically generated e-mail. To

Re: Review Request 50123: Added GPU scheduler for docker containerizer process.

> On 八月 25, 2016, 12:32 a.m., Kevin Klues wrote: > > src/slave/containerizer/docker.hpp, lines 20-25 > > > > > > What does adding all of these headers have to do with this patch? They > > may be needed, but I don't s

Re: Review Request 51320: Implemented the LaunchGroup Offer::Operation in the master.

> On 八月 23, 2016, 11:48 a.m., Guangya Liu wrote: > > src/master/master.cpp, line 93 > > <https://reviews.apache.org/r/51320/diff/1/?file=1481651#file1481651line93> > > > > I'd prefer that we move this under `using std::xxx`. > > > >

Re: Review Request 45081: Excluded reserved resources when got nonRevocable resources in stage 1.

> On 八月 11, 2016, 10:16 a.m., Michael Park wrote: > > Hi Klaus, could you explain what the motivation is for this patch? > > Currently, your analysis seems correct that reserved resources are always > > non-revocable. > > However, the current code seems that it'll be more future-proof. > > That

Re: Review Request 49617: Add benchmark for failover of many frameworks.

ll of the `allocations` are available when `recoverResources`? Do we need a `Clock::settle();` after all `suppressOffers` finshed in line 4090? - Guangya Liu On 八月 17, 2016, 2:26 a.m., Jacob Janco wrote: > > --- > This is an autom

Re: Review Request 49617: Add benchmark for failover of many frameworks.

> On 八月 18, 2016, 2:31 a.m., Guangya Liu wrote: > > src/tests/hierarchical_allocator_tests.cpp, line 4109 > > <https://reviews.apache.org/r/49617/diff/7/?file=1476268#file1476268line4109> > > > > What about update the message as > > > >

<    1   2   3   4   5   6   7   8   9   10   >