Re: PCS Management Counters
Vivek,
Counting that way doesn't work either. Picture a long
packet with an error near the beginning and another
error near the end. The counter would increment twice
for that packet.
As I tried to say earlier, this counter is a diagnostic
tool only. It is not a solution. A test person can use
it to monitor the quality of the link by reading it
continuously and tracking the rate of errors.
Most MACs already count the number of packets discarded.
Ben
> Vivek Mishra wrote:
>
> Ben,
>
> Incrementing the ERRORED FRAME COUNT when the
> state machine enters the RX_E state
> "only" for the first time would give the
> count of the number of packets thrown
> away because of ERROR column being there.
>
> Don't you think so this will be a better
> statistic to have about the quality of the
> link - the number of packets that had
> to be thrown away - rather than just a count
> which over a period of time is just a number?
>
> cheers,
> vivek
>
> -----Original Message-----
> From: Ben Brown [mailto:bbrown@xxxxxxxx]
> Sent: Monday, January 08, 2001 7:49 PM
> To: 802.3ae
> Subject: Re: PCS Management Counters
>
> Vivek,
>
> During the discussion of adding these counters in the Clause 49
> breakout session in Tampa, it was suggested that these counters
> could be used to give a user a feel for the quality of the link.
>
> It isn't possible to correlate the value in errored_frame_count
> with the number of packets lost on a link or the number of bad
> packets counted by a MAC. However it is often useful to have a
> link quality measure such as this counter available for diagnostic
> purposes.
>
> Enjoy,
> Ben
>
> > Vivek Mishra wrote:
> >
> > Ben,
> > Thanks for your reply.
> >
> > Where will these counters be used?
> > I mean do these counters give some kind of
> > count which tells the number of packets that
> > will be thrown away by the MAC because of
> > "ERROR" column?
> >
> > If the counter gives the number of ERROR columns
> > inserted by PHY how can somebody use
> > it for statistics?
> >
> > vivek
> >
> > -----Original Message-----
> > From: Ben Brown [mailto:bbrown@xxxxxxxx]
> > Sent: Friday, January 05, 2001 5:38 PM
> > To: 802.3ae
> > Subject: Re: PCS Management Counters
> >
> > Vivek,
> >
> > Since, according to the state machine conventions and figure 49-14,
> > the state is actually re-entered on every clock edge, the counter
> > would be incremented twice (in your case) or once for every clock
> > cycle spent in that state (in general).
> >
> > Ben
> >
> > > Vivek Mishra wrote:
> > >
> > > Hi all,
> > >
> > > Just a small clarification.....
> > >
> > > The counter "ERRORED FRAME COUNTER" in PCS management counters is
> > > said to be incrementing when the "RX_E" state is entered.
> > >
> > > Lets say that the state machine is in "RX_E" state for two clocks.
>
> > >
> > > Will the counter increment by two or will it increment
> > > by one since the state was entered only once?
> > >
> > > Regards,
> > > Vivek
> >
> > --
> > -----------------------------------------
> > Benjamin Brown
> > AMCC
> > 2 Commerce Park West
> > Suite 104
> > Bedford NH 03110
> > 603-641-9837 - Work
> > 603-491-0296 - Cell
> > 603-626-7455 - Fax
> > 603-798-4115 - Home Office
> > bbrown@xxxxxxxx
> > -----------------------------------------
>
> --
> -----------------------------------------
> Benjamin Brown
> AMCC
> 2 Commerce Park West
> Suite 104
> Bedford NH 03110
> 603-641-9837 - Work
> 603-491-0296 - Cell
> 603-626-7455 - Fax
> 603-798-4115 - Home Office
> bbrown@xxxxxxxx
> -----------------------------------------
--
-----------------------------------------
Benjamin Brown
AMCC
2 Commerce Park West
Suite 104
Bedford NH 03110
603-641-9837 - Work
603-491-0296 - Cell
603-626-7455 - Fax
603-798-4115 - Home Office
bbrown@xxxxxxxx
-----------------------------------------