> On June 2, 2015, 12:57 a.m., Chi Zhang wrote: > > Cong, is this the patch the requires code from 3.2.26 for the first time? > > If so, should we add a compile-time flag to config.ac? That way Paul can > > safely back off the code for the "stat string conversion" from the other > > patch you commented on.
No, https://reviews.apache.org/r/31503/ this one should be the first commit requires 3.2.26. - Cong ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/31505/#review86133 ----------------------------------------------------------- On June 1, 2015, 10:45 p.m., Cong Wang wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/31505/ > ----------------------------------------------------------- > > (Updated June 1, 2015, 10:45 p.m.) > > > Review request for mesos, Chi Zhang, Ian Downes, and Jie Yu. > > > Bugs: MESOS-2422 > https://issues.apache.org/jira/browse/MESOS-2422 > > > Repository: mesos > > > Description > ------- > > Currently we do nothing on the host egress side. By default, kernel uses its > own hash function to classify the packets to different TX queues (if the > hardware interface supports multiqueue). So packets coming out of different > containers could end up queueing in the same TX queue, in this case we saw > buffer bloat on some TX queue caused packet drops. > > We need to isolation the egress traffic so that containers will not have > interference with each other. The number of hardware TX queues is limited by > hardware interface, usually not enough to map our container in 1:1 way, > therefore we need some software solution. We choose fq_codel and use tc > filters to classify packets coming out of different containers to different > fq_codel flows, and the codel algorithm on each flow could also help us to > reduce the buffer bloat. Note when the packets leave fq_codel, they still > share the physical TX queue(s), this is however (almost) beyond what we can > control, we have to rely on the kernel behavior. > > TODO: get some performance numbers > > > Diffs > ----- > > src/linux/routing/filter/basic.cpp 4ce8acb7040f2ed8ec3834dd7702a20f5316c20f > src/linux/routing/filter/filter.hpp > aaca57fbe80e3ffa3dd2c2bbed93849013b7a382 > src/linux/routing/filter/icmp.cpp 76877fb94a1c1e79133b8975419d2ea7d0300650 > src/linux/routing/filter/ip.hpp 9645f9488938c55fec253b36d9fa30eae72a8ca2 > src/linux/routing/filter/ip.cpp 0f3b856bd04f6a881e35452a24399715cd8a174f > src/slave/containerizer/isolators/network/port_mapping.hpp > 6579286b779882bec493a8e0f10486adc316dc6e > src/slave/containerizer/isolators/network/port_mapping.cpp > 843e52d6f9923d0ee0a0297cd5c464b8b72f5de3 > > Diff: https://reviews.apache.org/r/31505/diff/ > > > Testing > ------- > > Manually start two mesos containers with netperf running side. > > > Thanks, > > Cong Wang > >