----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/11122/#review20717 -----------------------------------------------------------
src/slave/slave.cpp <https://reviews.apache.org/r/11122/#comment42781> Why did you have to do this? With cgroups, the expectation is that the resources a slave sends out are usable by the frameworks because we do isolation. Giving an offer and then rejecting seems wrong to me. I think the right way to attack highly loaded slaves is to fix cgroups, if there is a bug there. Makes sense? - Vinod Kone On May 14, 2013, 5:57 p.m., Brenden Matthews wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/11122/ > ----------------------------------------------------------- > > (Updated May 14, 2013, 5:57 p.m.) > > > Review request for mesos. > > > Description > ------- > > From 01f88dc840e4beda78933d1b63eba5d2b6556573 Mon Sep 17 00:00:00 2001 > From: Brenden Matthews <[email protected]> > Date: Mon, 29 Apr 2013 11:35:53 -0700 > Subject: [PATCH 15/24] Slave feature: maximum system load. > > When the load exceeds a specified value, don't accept tasks. Some nodes > may become unstable under excessive load (i.e., heavy disk I/O), and > this helps prevent the assigning of further tasks to busy slaves. > > Review: https://reviews.apache.org/r/11122 > --- > src/slave/flags.hpp | 11 ++++++++++- > src/slave/slave.cpp | 43 +++++++++++++++++++++++++++++++++++++++++++ > 2 files changed, 53 insertions(+), 1 deletion(-) > > > Diffs > ----- > > src/slave/flags.hpp f3cbe3d5b633e149a349c4ec9784ac5ebfbd4325 > src/slave/slave.cpp d3126cc072cbf17cbc5cc18be4bfa4840886fb16 > > Diff: https://reviews.apache.org/r/11122/diff/ > > > Testing > ------- > > Used in production at airbnb. > > > Thanks, > > Brenden Matthews > >
