> On Feb. 3, 2017, 1:34 p.m., Alexander Rukletsov wrote: > > src/tests/health_check_tests.cpp, lines 2118-2119 > > <https://reviews.apache.org/r/55901/diff/3/?file=1622284#file1622284line2118> > > > > Why do we need to expilictly create `containerizer`?
Because the implicit `containerizer` is started in local mode. `LaunchNestedContainerSession` (used by cmd health checks) tries to start a IO switchboard, which doesn't work in local mode yet. > On Feb. 3, 2017, 1:34 p.m., Alexander Rukletsov wrote: > > src/checks/health_checker.hpp, line 207 > > <https://reviews.apache.org/r/55901/diff/3/?file=1622281#file1622281line207> > > > > How about calling it `commandCheckViaAgent`? Good idea, done. > On Feb. 3, 2017, 1:34 p.m., Alexander Rukletsov wrote: > > src/checks/health_checker.cpp, line 130 > > <https://reviews.apache.org/r/55901/diff/3/?file=1622282#file1622282line130> > > > > s/createRequest/createV1AgentAPIRequest? > > remove `inline` > > > > This looks like a utility function not really related to the health > > checker library. Maybe put it into "src/slave/api_utils.{hpp|cpp}"? I renamed the function. Vinod suggested to make it `inline` what's wrong with that? This function as-is is specific to how we perform requests in this file (`keep_alive = false`, protobuf serialization). Since it is used only in two places, I'm tempted to removed it and put the code inline in the corresponding methods. > On Feb. 3, 2017, 1:34 p.m., Alexander Rukletsov wrote: > > src/tests/health_check_tests.cpp, line 2107 > > <https://reviews.apache.org/r/55901/diff/3/?file=1622284#file1622284line2107> > > > > Why not spelling out `cmd`? Renamed. > On Feb. 3, 2017, 1:34 p.m., Alexander Rukletsov wrote: > > src/tests/health_check_tests.cpp, lines 2154-2155 > > <https://reviews.apache.org/r/55901/diff/3/?file=1622284#file1622284line2154> > > > > Formatting. > > > > Also it is probably a good idea to explicitly say you don't care about > > further updates. Fixed the formatting. We don't have such a comment in most of the other tests in this file. - Gastón ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/55901/#review164117 ----------------------------------------------------------- On Feb. 3, 2017, 5:15 p.m., Gastón Kleiman wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/55901/ > ----------------------------------------------------------- > > (Updated Feb. 3, 2017, 5:15 p.m.) > > > Review request for mesos, Alexander Rukletsov, Anand Mazumdar, haosdent > huang, and Vinod Kone. > > > Bugs: MESOS-6280 > https://issues.apache.org/jira/browse/MESOS-6280 > > > Repository: mesos > > > Description > ------- > > Added support for command health checks to the default executor. > > > Diffs > ----- > > src/checks/health_checker.hpp 95da1ff7dd6b222a93076633eb3757ec9aa43cf6 > src/checks/health_checker.cpp 58380dc18896f659aa9c4fb4bb567a55bba97f6b > src/launcher/default_executor.cpp 97eee05cac8cb1f62d43e2aecc08a8e54e49eac3 > src/tests/health_check_tests.cpp 8418cd91484fd26734de16255b37f3ebf574f5eb > > Diff: https://reviews.apache.org/r/55901/diff/ > > > Testing > ------- > > Introduced a new test: `HealthCheckTest.DefaultExecutorCmdHealthCheck`. It > passes on Linux, but not on macOS. > > > Thanks, > > Gastón Kleiman > >
