Hi, group.
I just want to share something I encountered while doing FR labs. This
stemmed
from doing Lab 13 of Solie's book. Anyway, if you don't have the book, the
scenario is basic map statements for spoke1 and spoke 2 running on Physical
Interfaces pointing to the Hub and InARP and ARP Frame disabled. The Hub has
a
multipoint with mapping to each spoke. There is EIGRP running on all three.
My test's goal is to verify if indeed a Layer 2 mapping (static or dynamic)
must exist before Layer 3 reachability is achieved in Frame Relay. What I
found was this...
1. For spoke 1, all I needed was a map to the hub and have a routing protocol
advertise spoke 2's subnet(s); as long as spoke 2 has a mapping to spoke 1
also configured. This scenario gives me connectivity from spoke 1 to spoke
2's
subnet(s) even though I have no mapping for spoke 2's IP.
4500-E#sh frame map
Serial2 (up): ip 10.10.1.9 dlci 111(0x6F,0x18F0), static,
broadcast,
CISCO, status defined, active
4500-E#ping 10.10.5.5
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.10.5.5, timeout is 2 seconds:
!!!!
00:17:26: Serial2(i): dlci 111(0x18F1), pkt type 0x800, datagramsize 64
00:17:27: Serial2(o): dlci 111(0x18F1), pkt type 0x800(IP), datagramsize 104
00:17:27: Serial2(i): dlci 111(0x18F1), pkt type 0x800, datagramsize 60
00:17:27: Serial2(i): dlci 111(0x18F1), pkt type 0x800, datagramsize 104
00:17:27: Serial2(o): dlci 111(0x18F1), pkt type 0x800(IP), datagramsize 104
00:17:27: Serial2(i): dlci 111(0x18F1), pkt type 0x800, datagramsize 104
00:17:27: Serial2(o): dlci 111(0x18F1), pkt type 0x800(IP), datagramsize 104
00:17:27: Serial2(i): dlci 111(0x18F1), pkt type 0x800, datagramsize 104
00:17:27: Serial2(o): dlci 111(0x18F1), pkt type 0x800(IP), datagramsize 104!
Success rate is 100 percent (5/5), round-trip min/avg/max = 52/56/64 ms
4500-E#
00:17:27: Serial2(i): dlci 111(0x18F1), pkt type 0x800, datagramsize 104
00:17:27: Serial2(o): dlci 111(0x18F1), pkt type 0x800(IP), datagramsize 104
00:17:27: Serial2(i): dlci 111(0x18F1), pkt type 0x800, datagramsize 104
2. But with the same scenario, spoke 2 might have a mapping to spoke 1 and
have spoke1's subnet(s) in its routing table, yet there is no reachability at
all, as this shows...
4500-F(config)#int s2
4500-F(config-if)#frame map ip 10.10.1.10 121 broadcast
4500-F(config-if)#^Z
00:27:11: %SYS-5-CONFIG_I: Configured from console by console
4500-F#sh frame map
Serial2 (up): ip 10.10.1.9 dlci 121(0x79,0x1C90), static,
broadcast,
CISCO, status defined, active
Serial2 (up): ip 10.10.1.10 dlci 121(0x79,0x1C90), static,
broadcast,
CISCO, status defined, active
4500-F#ping 10.10.3.3
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.10.3.3, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)
In case someone wants to know, this is the version loaded in both spokes...
c4500-a3jk8s-mz.122-5.bin
Also, I think the only reason why the author explicitly entered the commands
"no frame inverse" and "no arp frame" was to prevent the spokes from
obtaining
DYNAMIC mapping to each other if you FORGOT to enter static mappings to each
other, since there's nowhere in the lab where you remove the previously
configured frame route for both spokes at the FRAME-SWITCH.
N.B. Reinitializing the spoke's serial interfaces made no difference. Same
holds true for "clear frame inarp".
N.N.B. Enabling/disabling ARP FRAME and FRAME INVERSE-ARP made no difference,
as obviously the static entries are preferred.
Any comments on this one-way street behavior will be greatly appreciated.
Thanks.
Elmer
"What problem are you trying to study, group?"
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=46198&t=46198
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
