
377
Cross-Platform Release Notes for Cisco IOS Release 12.0S
OL-1617-14 Rev. Q0
Caveats
Resolved Caveats—Cisco IOS Release 12.0(33)S7
http://www.cisco.com/en/US/docs/ios/12_3t/12_3t4/feature/guide/gtrtlimt.html
Further Problem Description: Cisco IOS Software releases that have the fix for this Cisco bug ID,
have a behavior change for mode 7 private mode packets.
Cisco IOS Software release with the fix for this Cisco bug ID, will not process NTP mode 7 packets,
and will display a message “NTP: Receive: dropping message: Received NTP private mode packet.
7” if debugs for NTP are enabled.
To have Cisco IOS Software process mode 7 packets, the CLI command ntp allow mode private
should be configured. This is disabled by default.
• CSCtd90953
Symptoms: A ping delay is observed on E5 interfaces.
Conditions: This symptom is observed after a CSC switchover is performed.
Workaround: Reload the linecard.
• CSCtd93825
Symptoms: When the E4+ linecard is inserted into a blank slot of a Cisco 12000 series router, the
MAC addresses of that linecard becomes 0000.0000.0000.
Conditions: This symptom does not occur if the system boots with the linecard; this symptom occurs
only when the E4+ linecard is inserted into a blank slot after booting is complete.
Workaround: Reload the system.
• CSCte45025
Symptoms: The main Gigabit interface flaps.
Conditions: This symptom occurs when a Gigabit subinterface with an IPv6 address is deleted.
Workaround: There is no workaround.
• CSCte83282
Symptoms: Submillisecond drops of ATM cells going over AToM pseudo-wire MPLS are observed.
Conditions: This symptom occurs whenever the TE tunnel used by the pseudo-wire is reoptimized.
Workaround: If possible, avoid the trigger for reoptimization (such as auto-bw or loose path).
• CSCte83622
Symptoms: High CPU utilization is seen on Cisco IOS Release 12.0(33)S5 code due to “RIP Send”
and “BGP Scanner.”
Conditions: RIP in Cisco IOS Release 12.0(33)S does not have a separate database about the routes.
When RIP has to send update packets, it will scan the whole routing table, validate each prefix and
create the update packets and send out. Above event will have to happen every 15 seconds as per the
timers configuration.
With a network where there are 150,000 routes learned by BGP, it is quite possible the routing table
scan during update generation can be CPU intensive.
We will see the problem of continuous rip updates and high CPU until an RIP Send process.
Please note that a Supernet prefix needs to be installed in the RIB, and RIP has to include them in
its updates. If it is originated by static or other routing protocols, we will see the symptom if they
are redistributed into RIP.
Comentarios a estos manuales