mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jie Yu" <yujie....@gmail.com>
Subject Re: Review Request 34426: Report per-container metrics for network bandwidth throttling
Date Thu, 04 Jun 2015 16:39:01 GMT


> On June 3, 2015, 10:35 p.m., Jie Yu wrote:
> > src/linux/routing/queueing/internal.hpp, line 320
> > <https://reviews.apache.org/r/34426/diff/11/?file=977689#file977689line320>
> >
> >     Why <=?
> 
> Paul Brett wrote:
>     The definition for RTNL_TC_STATS_MAX comes from libnl and looks like this:
>     
>     enum rtnl_tc_stat {
>       RTNL_TC_PACKETS,  /**< Number of packets seen */
>       RTNL_TC_BYTES,    /**< Total bytes seen */
>       RTNL_TC_RATE_BPS, /**< Current bits/s (rate estimator) */
>       RTNL_TC_RATE_PPS, /**< Current packet/s (rate estimator) */
>       RTNL_TC_QLEN,   /**< Current queue length */
>       RTNL_TC_BACKLOG,  /**< Current backlog length */
>       RTNL_TC_DROPS,    /**< Total number of packets dropped */
>       RTNL_TC_REQUEUES, /**< Total number of requeues */
>       RTNL_TC_OVERLIMITS, /**< Total number of overlimits */
>       __RTNL_TC_STATS_MAX,
>     };
>     #define RTNL_TC_STATS_MAX (__RTNL_TC_STATS_MAX - 1)
>     
>     So the range is from 0 to RTNL_TC_STATS_MAX inclusive.
> 
> Cong Wang wrote:
>     This looks like a libnl bug for me, however it is too late to fix it...

OK, I added a NOTE for you because this is very unintuitive.


> On June 3, 2015, 10:35 p.m., Jie Yu wrote:
> > src/linux/routing/queueing/statistics.hpp, line 23
> > <https://reviews.apache.org/r/34426/diff/11/?file=977690#file977690line23>
> >
> >     Should we get rid of 'queueing' here because this can be used by filters as
well?
> 
> Paul Brett wrote:
>     It should work on filters but we have no use case and no testing to know if it works.
 Lets leave it here until that changes.

SGTM.


- Jie


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/34426/#review86493
-----------------------------------------------------------


On June 4, 2015, 1:19 a.m., Paul Brett wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/34426/
> -----------------------------------------------------------
> 
> (Updated June 4, 2015, 1:19 a.m.)
> 
> 
> Review request for mesos, Chi Zhang, Ian Downes, Jie Yu, and Cong Wang.
> 
> 
> Bugs: MESOS-2665
>     https://issues.apache.org/jira/browse/MESOS-2665
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Report the network statistics from libnl
> 
> 
> Diffs
> -----
> 
>   src/Makefile.am 66030c4b211ea61e97e62c35ec1821e0958f9787 
>   src/linux/routing/queueing/fq_codel.hpp 49c8df626addcfc841132f821b8697b5e6e3b341 
>   src/linux/routing/queueing/fq_codel.cpp 976c6a7a11d39609e6597daeb1a74c93ac62e4bf 
>   src/linux/routing/queueing/ingress.hpp 2b7e1d3d1c02120195b6697320fb46f085a1ef92 
>   src/linux/routing/queueing/ingress.cpp e96f547200dc4ad5b2e33fd1ffbd7fb92b955a46 
>   src/linux/routing/queueing/internal.hpp 3713f6a4d3f9a44b142b8150e85cf911e53e34e9 
>   src/linux/routing/queueing/statistics.hpp PRE-CREATION 
> 
> Diff: https://reviews.apache.org/r/34426/diff/
> 
> 
> Testing
> -------
> 
> make check
> 
> 
> Thanks,
> 
> Paul Brett
> 
>


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