> On Jan. 18, 2017, 1:06 a.m., Joseph Wu wrote: > > 3rdparty/libprocess/src/process.cpp, line 1328 > > <https://reviews.apache.org/r/55024/diff/2/?file=1605541#file1605541line1328> > > > > At the moment, I would not expect this to ever fail, > > `EXIT(EXIT_FAILURE)` is preferable. If it does, it seems like a severe > > enough error (as the docs suggest most errors are due to programmer error). > > > > See: > > https://msdn.microsoft.com/en-us/library/windows/desktop/ms741549(v=vs.85).aspx > > Alex Clemmer wrote: > The reason I decided to `LOG(ERROR)` is because there didn't seem to be > any scenarios where we would encounter errors because we haven't shut down > the socket stack. We can change it if you feel strongly about it, though. Do > you think there is a strong possibility of an error condition that I'm > missing? > > Joseph Wu wrote: > We generally err on the side of failing-fast. So if we don't expect > something to occur, we first place a `CHECK` or `EXIT`. If that turns out to > be incorrect later, we consider relaxing the code at that point. > > This usually gives more visibility into problems, as people are more > likely to notice a crash than an error log.
I agree about fail-fast, which is good distributed systems hygiene. It just seems to me that the dispose path is basically never the cause of service errors. But, anyway, let's change it. I just wanted to double-check there wasn't a specific error case I was missing. - Alex ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/55024/#review162008 ----------------------------------------------------------- On Jan. 18, 2017, 5:41 p.m., Alex Clemmer wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/55024/ > ----------------------------------------------------------- > > (Updated Jan. 18, 2017, 5:41 p.m.) > > > Review request for mesos, Andrew Schwartzmeyer, Daniel Pravat, John Kordich, > and Joseph Wu. > > > Repository: mesos > > > Description > ------- > > Currently libprocess will attempt to use sockets without initializing > the socket stack on Windows. This commit will resolving this problem by > causing `process::initialize` to initialize the socket stack. > > > Diffs > ----- > > 3rdparty/libprocess/include/process/process.hpp > b118f1a2bf5aac12b53ae204253b88c9b1c65f46 > 3rdparty/libprocess/src/process.cpp > f475fe78f801924f70f51fdc4ab190c2dbecd656 > > Diff: https://reviews.apache.org/r/55024/diff/ > > > Testing > ------- > > > Thanks, > > Alex Clemmer > >