Gents & Ladies,
I am trying to replicate a scenario I was given in a lab and cannot
get the same results when I tried it on dynamips. The situation
involves using "ntp access-group serve-only ACL#" feature on R6.
-R6 is pulling timing from BB1. BB1 is the clock source
-R6 can only allow R3 to pull timing from it
-R3 must pull timing from R6
My answer:
R6#
ntp server 54.1.2.254 (BB1)
ntp access-group serve-only 1
ip access-list 1 permit 150.1.3.3
R3#
ntp server 150.1.6.6
ntp source loopback0
Correct solution:
R6#
ntp server 54.1.2.254 (BB1)
ntp access-group serve-only 1
ntp access-group peer 2
ip access-list 1 permit 150.1.3.3
ip access-list 2 permit 54.1.2.254
R3#
ntp server 150.1.6.6
ntp source loopback0
Explanation given:
NTP access group options are scanned in the following order
from least restrictive to most restrictive:
- peer
- serve
- serve-only
- query-only
Access is granted for the first match that is found. If no access groups
are specified, all access is granted to all sources. If any access
groups
are specified, only the specified access is granted.
If only the serve option is configured your router is not allowed to
synchronize to the remote system.
My QUESTION:
Is the "ntp acess-group peer #" really required as seen in the correct
solution? When I simulated this behavior, R3 was able to successfully
pull timing from R6 however any devices not in ACL#1 were not able to
pull the timing. The access-group peer was not required on R6, it
kept its timing from BB1 no problems. I waited a few hours.
What am I missing? Is this IOS behavior dependent?
Thanks,
Gary
Blogs and organic groups at http://www.ccie.net
Received on Fri Jul 24 2009 - 16:37:34 ART
This archive was generated by hypermail 2.2.0 : Sat Aug 01 2009 - 13:10:23 ART