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 31502: (1/4) Add fq_codel qdisc for egress
Date Mon, 02 Mar 2015 17:54:05 GMT


> On Feb. 26, 2015, 11:19 p.m., Dominic Hamon wrote:
> > src/linux/routing/queueing/fq_codel.hpp, line 33
> > <https://reviews.apache.org/r/31502/diff/1/?file=879067#file879067line33>
> >
> >     static is a better fit than extern for this. also does it actually have to be
in the header?

fq_codel::HANDLE is used by port_mapping.c in the last patch.


- Cong


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/31502/#review74389
-----------------------------------------------------------


On Feb. 27, 2015, 9:38 p.m., Cong Wang wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/31502/
> -----------------------------------------------------------
> 
> (Updated Feb. 27, 2015, 9:38 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
> -------
> 
> Fq_codel qdisc is needed for egress flow classification. Internally fq_codel defines
its own flows, and it allows us to specifiy the rule to classify the flows with tc filters
attached on it. By default it has 1024 flows, much more than the number of containers we could
create on a single host.
> 
> 
> Diffs
> -----
> 
>   src/Makefile.am 17d0d7aa7361c3a373f6863d36b0a4767f5c05c4 
>   src/linux/routing/queueing/fq_codel.hpp PRE-CREATION 
>   src/linux/routing/queueing/fq_codel.cpp PRE-CREATION 
>   src/tests/routing_tests.cpp 3cda6ab8c1ad24e4b7d0b9aeda2abc595fc857a5 
> 
> Diff: https://reviews.apache.org/r/31502/diff/
> 
> 
> Testing
> -------
> 
> Run the testcase.
> 
> 
> Thanks,
> 
> Cong Wang
> 
>


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