mesos-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Cong Wang" <cw...@twopensource.com>
Subject Re: Review Request 31505: (4/4) Add flow classifiers for fq_codel on egress
Date Mon, 02 Mar 2015 17:33:07 GMT


> On Feb. 27, 2015, 7:50 a.m., Ian Downes wrote:
> > src/slave/containerizer/isolators/network/port_mapping.cpp, lines 1828-1847
> > <https://reviews.apache.org/r/31505/diff/1/?file=879084#file879084line1828>
> >
> >     I'm confused - do these need to be configured for each container? I thought
all (host and container) icmp traffic was on the same flow?

All ICMP traffic is on the same flow, this filter should only be created once because it is
inside the "if (targets.size() == 1)" check.


- Cong


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31505/#review74457
-----------------------------------------------------------


On March 2, 2015, 5:21 p.m., Cong Wang wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31505/
> -----------------------------------------------------------
> 
> (Updated March 2, 2015, 5:21 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/slave/containerizer/isolators/network/port_mapping.hpp 8443097b2c79fef5ae0e23a3fb815ffec0318a93

>   src/slave/containerizer/isolators/network/port_mapping.cpp 5227987cdb7b904c2f4bb2bdf5c5d705a435010d

> 
> Diff: https://reviews.apache.org/r/31505/diff/
> 
> 
> Testing
> -------
> 
> Manually start two mesos containers with netperf running side.
> 
> 
> Thanks,
> 
> Cong Wang
> 
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message