[nsp] Routing Problem: I am not sure where to begin.
Jim Devane
jim at powerpulse.cc
Tue Dec 16 12:13:11 EST 2003
Jack,
Thank you again.
Actually, I threw the GSR into Unidirectional to match the ADM.
Routers, as an IP device, cannot provide true Unidrectional protection.
Routers will not known a path to a source address out on interface and rcv
traffic from it on another. It would drive CEF mad.
Routers much prefer bidirectional APS. As such, "router" unidirectional
mode, performs 2 switches to, in essence, spoof Uni mode.
If the AIS is rcv'd it will send an AIS back upstream to force both Tx and
Rx over to the "protect" With bi-directional, of course everything is moved
on the the original receipt of the AIS.
Unfortunately, the 1+1 protection provided by the ADM is UPSR (uni) so I put
in Uni for testing to make sure the K2 bytes match 0x04
Your comments about voice vs data make complete sense to me, but
unfortunately it appears the Lucent has a limitation on how it can deliver 1
+ 1 APS redundancy. ( or so I am told )
Incidentally, the uni directional in the output is there b/c of the APS
mis-match. If there is a mis-match RCV K2 are 0x04 and Xmit are 0X05 the
router will go into Uni to compensate.
Even so, I do not believe this to be an APS/SONET issue as that is the one
thing that performs consistently and deterministically. It is the L3 routing
that seem to lag behind, but I cannot figure out why.
I am not sure what my next step is. But I very much appreciate your
suggestions/comments.
Thanks again,
Jim
-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of
Jack.W.Parks at alltel.com
Sent: Monday, December 15, 2003 6:50 PM
To: cisco-nsp at puck.nether.net
Subject: RE: [nsp] Routing Problem: I am not sure where to begin.
I think I just realized something. You are using APS in unidirectional
mode. This works great for voice, but horrible for data. In effect,
you might have Transmit on your working and Receive on your protect.
>From your configuration "provisioned bidirectional":
POS8/1 APS Group 20: protect channel 0 (inactive)
unidirectional (provisioned bidirectional), revertive (1 min)
We have experienced this problem ourselves. Being a Telco, our
transport guys provisioned our SONET circuits APS unidirectional (for
voice). For data you want APS bidirectional, so your transmit and
receive are on the same fiber pair, remove that line in your config. It
seems you are provisioned for bidirectional support and that is what you
should to use. I'm not a SONET transport engineer, but maybe someone on
the list knows the difference between uni- and bi-directional APS
support (other than the obvious stated in the name).
Jack
-----Original Message-----
From: Jim Devane [mailto:jim at powerpulse.cc]
Sent: Monday, December 15, 2003 7:17 PM
To: 'Gert Doering'; 'Ryburn, Justin'
Cc: Parks, Jack W; cisco-nsp at puck.nether.net
Subject: RE: [nsp] Routing Problem: I am not sure where to begin.
Gert,
Thank you as well for replying.
Yes, HSRP will not work here, I agree.
I also agree that it *should* work. But dang it, I can't figure out why
it is not...especially since the SONET framing is correct and passing
the correct L1/L2 info between the routers.
(see below taken during an APS forced switch)
I did only check CEF adjacency on the RP itself. In my next window I
will check the CEF adj on the LC itself and see if there is a
difference.
But other than something weird in my CEF table I cannot think of why
packets are not going through.
Any ideas?
Thanks again for your response...
Jim
interface POS8/1
description working int
ip address 66.209.x.x 255.255.255.252
no ip directed-broadcast
load-interval 30
no keepalive
crc 16
pos ais-shut
aps group 20
aps unidirectional
aps working 1
end
sh aps
POS8/1 APS Group 20: working channel 1 (active)
PGP timers (from protect): hello time=1; hold time=3
SONET framing; SONET APS signalling by default
Protect at 66.209.X.X
Remote APS configuration: (null)
interface POS8/1
description protect int
ip address 66.209.x.x 255.255.255.252
no ip directed-broadcast
load-interval 30
no keepalive
crc 16
pos ais-shut
aps group 20
aps unidirectional
aps revert 1
aps working 1
end
sh aps
POS8/1 APS Group 20: protect channel 0 (inactive)
unidirectional (provisioned bidirectional), revertive (1 min)
PGP timers (default): hello time=1; hold time=3
state: (Disabled)
authentication = (default)
PGP versions (native/negotiated): 2/2
SONET framing; SONET APS signalling by default
Received K1K2: 0x00 0x04
No Request (Null)
Transmitted K1K2: 0x00 0x04
No Request (Null)
Working channel 1 at 66.209.X.X (Enabled)
Remote APS configuration: (null)
Reflected local configuration: (null)
FORENSIC:
#sh controllers pos8/1
POS8/1
SECTION
LOF = 6 LOS = 6 BIP(B1) =
131811
LINE
AIS = 6 RDI = 15 FEBE = 980 BIP(B2) = 456
PATH
AIS = 24 RDI = 42 FEBE = 8824 BIP(B3) = 869
LOP = 0 NEWPTR = 46 PSE = 4 NSE = 8
Active Defects: None
Active Alarms: None
Alarm reporting enabled for: SF SLOS SLOF B1-TCA B2-TCA PLOP B3-TCA
Framing: SONET
APS
protect (inactive)
COAPS = 74 PSBF = 0
State: PSBF_state = False
ais_shut = TRUE
Rx(K1/K2): 00/04 Tx(K1/K2): 00/04
Signalling protocol: SONET APS by default
Rx Synchronization Status S1 = 0x00
S1S0 = 00, C2 = CF
Remote aps status (none); Reflected local aps status (none) CLOCK
RECOVERY
RDOOL = 0
State: RDOOL_state = False
PATH TRACE BUFFER : STABLE
Remote hostname : las-core02
Remote interface: POS6/0
Remote IP addr : 66.209.X.X
Remote Rx(K1/K2): 00/00 Tx(K1/K2): 00/00
BER thresholds: SF = 10e-3 SD = 10e-6
TCA thresholds: B1 = 10e-6 B2 = 10e-6 B3 = 10e-6
-----Original Message-----
From: Gert Doering [mailto:gert at greenie.muc.de]
Sent: Monday, December 15, 2003 1:57 PM
To: Ryburn, Justin
Cc: Jim Devane; Jack.W.Parks at alltel.com; cisco-nsp at puck.nether.net
Subject: Re: [nsp] Routing Problem: I am not sure where to begin.
Hi,
On Mon, Dec 15, 2003 at 04:41:54PM -0500, Ryburn, Justin wrote:
> No, I missed that there were two different routers on the GSR side.
> The only way I know of to do it in that setup is to do HSRP between
> the two GSRs and peer to the virtual address. I am not personally a
> big fan of HSRP but in theory that should work.
We're talking APS protection switching. There is no LAN here that you
could run HSRP on.
As for Jim's orignal question: I've never done POS APS, but I'd think it
*should* work. If the interface is properly downed when the circuit is
not active, and "up/up" when it's active, I can't see a specific reason
why using the same set of IP addresses shouldn't work.
It certainly would work that way if you have "manual failover" (two
serial interfaces on two neighbouring routers, configured identically,
and you move the WAN line from one to the other).
gert
--
USENET is *not* the non-clickable part of WWW!
//www.muc.de/~gert/
Gert Doering - Munich, Germany
gert at greenie.muc.de
fax: +49-89-35655025
gert at net.informatik.tu-muenchen.de
_______________________________________________
cisco-nsp mailing list cisco-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
More information about the cisco-nsp
mailing list