Filter
Top Products
11-7
Cisco WAN Manager User’s Guide
Version 10.5, Part Number 78-12945-01 Rev. D0, August 2003
Chapter 11 CWM to CWM Communications
Recovering From Degrade Mode
Note An Abnormal Exit happens when a CWM Gateway process is stopped through a non-graceful
shutdown. This includes when a process has been stopped with a process core dump command,
but does not include a power down.
Unexpected Exit of the Primary CWMGateway
The unexpected exit of the Primary CWMGateway will be detected by the Secondary CWMGateway
through the Orbix callback (CORBA::IT_IOCallback) mechanism, which is invoked when the CORBA
connection between them is broken (provided that there is IP connectivity between the Primary and the
Secondary). The CWMGateway process will automatically be restarted by watchdog up to 5 times. Once
the Primary is restarted it will notify all Secondaries. Upon receiving this event, a Secondary
CWMGateway will re-register with the Primary.
If all Secondaries did not receive notification of the Primary restart either due to the Primary CWM’s
failure to restart after five attempts, or because IP connectivity was lost between the Secondary and the
Primary, the Secondaries will continue to work as Secondaries with no Primary CWM present, in a
degraded mode of operation.
Unexpected Exit of the Secondary CWMGateway
The unexpected exit of a Secondary CWMGateway will be detected by the Primary CWMGateway
through the Orbix callback (CORBA::IT_IOCallback) mechanism. The Primary CWMGateway will
reassign the priority (decremented by 1) of all Secondary CWMGateways whose priority number was
higher than the priority number of the departed Secondary, and will remove the client entry
corresponding to this Secondary CWMGateway from its list.
Primary CWM has lost IP or physical connectivity to all Secondary CWMs
–
To recover from this degraded mode of operation, re-establish IP connectivity on the Primary
CWM and then all Secondary CWMs will automatically re-sync their user data tables with the
Primary CWM.
–
If you can’t re-establish IP connectivity on the Primary, then eliminate the downed Primary
from the CWM domain by following these steps:
Step 1 Manually remove the downed Primary from all the other CWM’s DomainGatewayList in the
CWMGateway.conf file
The following is an example of a CWMGateway.conf file with a DomainGatewayList consisting of
three CWM workstations named cwmws1, cwmws2, and cwmws3, with cwmws1 assigned as the
Primary and cwmws2 and cwmws3 assigned as Secondaries:
Debug level 2
DomainGatewayList cwmws1 cwmws2 cwmws3
ForcedSwitchOver
HeartBeatInterval 20