----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/40563/#review107452 -----------------------------------------------------------
src/slave/http.cpp (lines 318 - 322) <https://reviews.apache.org/r/40563/#comment166556> indent by 4 spaces. also, it's hard to tell that None() is the argument for statusUpdate() and not createStatusUpdate() src/slave/slave.cpp (line 2764) <https://reviews.apache.org/r/40563/#comment166559> this will be printed by itself on a new line. just do; LOG(INFO) << "Handling status update " << update << (pid.isSome() ? " from " + pid.get() : ""): src/slave/slave.cpp (lines 2792 - 2800) <https://reviews.apache.org/r/40563/#comment166565> ditto. separate LOG(WARNING) messages puts the text on separate lines. not what you want here. src/slave/slave.cpp <https://reviews.apache.org/r/40563/#comment166566> how about short-circuiting here when an ack doesn't need to be sent? //.... if (pid.isSome() && pid.get() == UPID()) { return; } src/slave/slave.cpp (line 3033) <https://reviews.apache.org/r/40563/#comment166567> s/status update/sending acknowledgement for status update/ src/slave/slave.cpp (lines 3042 - 3043) <https://reviews.apache.org/r/40563/#comment166568> LOG(WARNING) << "Ignoring sending acknowledgement for status update " << update << " of unknown executor"; - Vinod Kone On Nov. 21, 2015, 12:21 a.m., Anand Mazumdar wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/40563/ > ----------------------------------------------------------- > > (Updated Nov. 21, 2015, 12:21 a.m.) > > > Review request for mesos, Ben Mahler, Isabel Jimenez, Marco Massenzio, and > Vinod Kone. > > > Bugs: MESOS-3476 > https://issues.apache.org/jira/browse/MESOS-3476 > > > Repository: mesos > > > Description > ------- > > This change adds the ability for the agent to handle status updates from HTTP > based executors. Previously, the existing `statusUpdate` method used to > handle status updates sent from slave with `pid == UPID()`, sent from other > executors/on behalf of other executors with `pid == Some()`. This change > modifies the argument to be `Option<PID>`. This ensures that `pid == None()` > when set, the ACK is correctly routed to the corresponding HTTP based > executor. > > > Diffs > ----- > > src/slave/http.cpp ce48a0584ab18a8d95dd02619f62df18b2276639 > src/slave/slave.hpp ec2dfa99e6b553e2bcd82d12db915ae8625075a1 > src/slave/slave.cpp d1126f00d947fdb4823b0c495335b743254ac7ee > > Diff: https://reviews.apache.org/r/40563/diff/ > > > Testing > ------- > > make check + would add tests once the https://reviews.apache.org/r/39297/ > chain is reviewed. > > > Thanks, > > Anand Mazumdar > >
