Re: dev.bce.X.com_no_buffers increasing and packet loss

From: Pyun YongHyeon <pyunyh_at_gmail.com>
Date: Tue, 9 Mar 2010 12:49:27 -0800
On Tue, Mar 09, 2010 at 03:31:55PM +0200, Ian FREISLICH wrote:
> Pyun YongHyeon wrote:
> > On Mon, Mar 08, 2010 at 04:45:20PM +0200, Ian FREISLICH wrote:
> > > Pyun YongHyeon wrote:
> > > > On Fri, Mar 05, 2010 at 11:16:41PM +0200, Ian FREISLICH wrote:
> > > > > Pyun YongHyeon wrote:
> > > > > > Thanks for the info. Frankly, I have no idea how to explain the
> > > > > > issue given that you have no heavy load.
> > > > > 
> > > > > How many cores would be involved in handling the traffic and runnig
> > > > > PF rules on this machine?  There are 4x
> > > > > CPU: Quad-Core AMD Opteron(tm) Processor 8354 (2194.51-MHz K8-class CPU
> )
> > > > > In this server.  I'm also using carp extensively.
> > > > > 
> > > > 
> > > > pf(4) uses a single lock for processing, number of core would have
> > > > no much benefit.
> > > 
> > > What's interesting is the effect on CPU utilisation and interrupt
> > > generation that net.inet.ip.fastforwarding has:
> > > 
> > > net.inet.ip.fastforwarding=1
> > > interrupt rate is around 10000/s per bce interface
> > > cpu 8.0% interrupt
> > > 
> > 
> > Yes, this is one of intentional change of the patch. Stock bce(4)
> > seems to generate too much interrupts on BCM5709 so I rewrote
> > interrupt handling with the help of David. sysctl nodes are also
> > exported to control interrupt moderation so you can change them if
> > you want. Default value was tuned to generate interrupts less than
> > 10k per second and try to minimize latencies.
> 
> Can you explain the tunables please - I'm guessing it's these:
> 
> dev.bce.$i.tx_quick_cons_trip_int

This value controls the number of TX Quick BD Chain entries that
must be completed before a status block is generated during an
interrupt.

> dev.bce.$i.tx_quick_cons_trip

This value controls the number of TX Quick BD Chain entries that
must be completed before a status block is generated. Setting this
to 0 disables TX Quick BD Chain consumption from generating status
blocks.

> dev.bce.$i.tx_ticks_int

This value controls the number of 1us ticks that will be counted
for status block updates generated due to TX activity during
interrupt processing. Setting this value to 0 disables the TX
timer feature during interrupts.

> dev.bce.$i.tx_ticks

This value controls the number of 1us ticks that will be counted
before a status block update is generated due to TX activity.
Setting this value to 0 disables the TX timer feature.

> dev.bce.$i.rx_quick_cons_trip_int

This value controls the number of RX Quick BD entries that must be
completed before a status block is generated during interrupt
processing.

> dev.bce.$i.rx_quick_cons_trip

This value controls the number of RX Quick BD Chain entries that
must be completed before a status block is generated. Setting this
to 0 disables RX Event consumption from generating status blocks.

> dev.bce.$i.rx_ticks_int

This value controls the number of 1us ticks that will be counted
for status block updates generated due to RX activity during
interrupt processing. Setting this value to 0 disables the RX
timer feature during interrupts.

> dev.bce.$i.rx_ticks
> 

This value controls the number of 1us ticks that will be counted
before a status block update is generated due to RX activity.
Setting this value to 0 disables the RX timer feature.
Received on Tue Mar 09 2010 - 19:49:33 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:01 UTC