Advertisement
Q01850786
Q01856573
Q01794609
8. New Outstanding Issues
Q01612783
Q01896432
©2007-2008 Nortel Networks Limited
From 4.1.6 onwards, users can configure redundant sync interface. However, the
nd
default 2
sync interface and the management interface were configured to use the
same physical device. Due to this, management interface configuration is not getting
nd
applied if 2
sync is not configured. Also the management interface settings are over
nd
written with the 2
sync settings.
The issue is resolved in the release 4.2.3 by correcting the default devices for MGMT
device and Sync2.
During a policy push, acceleration would be turned OFF and turned ON again. This
is required to apply any changes in the rule base to the existing & new traffic.
As per the original design, during accel OFF, the particular SFD would be marked
for deletion which would mean that it cannot handle the traffic.
Hence when the SFA receives data that needs to be forwarded to a particular SFD
that's marked for deletion, the session would be re-binded to the next available SFD.
This behavior works well when there is a synchronization of that particular session
among the SFDs. But synchronization is not done for all the services, for eg. It is
recommended by NORTEL to turn OFF sync for services like http. In this case the
behavior stated would cause the termination of the sessions, which is not an
expected behavior.
The issue is resolved in4.2.3 by modifying the SFA behavior during policy push. SFA
would now forward the traffic to the same SFD which was handling the traffic
before "ACCEL OFF" so that the traffic won't be dropped due to non-
synchronization.
The System LED description as given in the Hardware Installation Manual is
incorrect.
The system status LEDs indicates the operational status of four fans, chassis, CPU
temperature, ambient temperature and the voltages (+5V and +12V).
The different glow states of LEDs are as follows:
• If the system is reset, the LED doesn't glow
• If system detects any problem with any of the CPU temperature, fan speed or
system voltages, the LED glows amber
• If the system is working, the LED glows in solid green
• If the system halts, LED flashes
TCP connections with TCP window-scaling option enabled, stall intermittently when
the session is started. The problem is found to be with CP firewall dropping the
initial packets due to TCP sequence error. These dropped packets when
retransmitted by server continue the service there after. The same problem doesn't
occur.
Hitless upgrade fails from the version 4.1.5_R65 to 4.2.3_R65. Normal upgrade
works fine in this case. So users upgrading from 4.1.5_R65 to 4.2.3_R65 are
requested to use the normal upgrade instead of Hitless Upgrade. This issue is
planned to fix by next release.
6
Advertisement
