Re: [EFM] On the Worship of Speed
How does it work on 128kb DSL links with PPOE today?
At 09:38 PM 1/23/2002 -0500, FEffenberger@xxxxxxxxxxxxxxxxx wrote:
>Link Speed is my shepherd, I shall not be in want.
>It makes me lie down in small buffers, It leads me beside quiet networks,
>It restores my cache. It guides me in paths of QoS for his name's sake.
>Even though I walk through the valley of the Loss of Packets, I will fear
>no Jitter, for Speed is with me; your MAC and your PHY, they comfort me.
>Psalm 23, King Geoff Version :-)
>In PON, the emulated point to point speed is configurable.
>It could be as small as 1 Mb/s. At such speeds, large packet
>induced jitter is indeed a problem. Consider a 1500 Byte
>packet. It will take 12 ms to deliver over a 1 Mb/s link.
>That will be an issue for TDM services.
>From: carlosal@xxxxxxxxxxxxxxxxxx [mailto:carlosal@xxxxxxxxxxxxxxxxxx]
>Sent: Wednesday, January 23, 2002 8:45 PM
>To: Chou, Joey
>Cc: Geoff Thompson; bob.barrett@xxxxxxxxxxxxxxx; mattsquire@xxxxxxx;
>Subject: [EFM] RE: [EFM-OAM] Performance monitoring, installation,
>trouble shoot ing.
>As you said:
> > I thought there should be a need for an OAM requirement to support
> > 802.1D user_priority bits to prevent voice packets from blocking
> > by huge Ethernet data packets.
>Assuming that the typical speed of the EFM link is going to be very high,
>this blocking effect is not really an issue, as it can be solved with good
>queue management. That's an implementation issue. For low speed links, this
>is a serious problem, because it takes a long time to transmit a big frame.
>The Frame Relay Forum took specific measures to avoid this situation; for
>examplo, big frames may be fragmented before transmission. It's better to
>avoid fragmentation, though, as it's a complex implementation that presents
>its own set of issues to be solved.
>That said, it's important to state that this is not enough to solve all
>issues related to the support of TDM emulation services over EFM. Speed is
>good, and may help a lot by avoiding the need for complex mechanisms such
>as fragmentation, but there are other issues to watch out, specially in the
>case of the P2MP implementation.