Customers can perform the following workaround, which is equivalent to upgrading to a fixed software release. Although the reload of the sysadmin VM is hitless, Cisco recommends performing this change during a maintenance window:
Step 1: Access the sysadmin VM:
RP/0/RSP1/CPU0:eXR#admin
Tue Mar 12 22:46:37.110 UTC root connected from 127.0.0.1 using console on host
Step 2: Run bash and edit the calvados_bootstrap.cfg file:
sysadmin-vm:0_RSP1:eXR# run bash
Tue Mar 12 22:46:44.224 UTC bash-4.3# vi /etc/init.d/calvados_bootstrap.cfg
Step 3: Edit the file by changing:
#CTRL_VRF=0
#MGMT_VRF=2
To:
CTRL_VRF=0
MGMT_VRF=2
Exit the file and save. In dual RSP/RP systems, the edit must be performed on both active and standby RSP/RPs.
Step 4: Reload the sysadmin VM (repeat for both in dual systems):
sysadmin-vm:0_RSP1:eXR# reload admin location 0/RSP1
Tue Mar 12 22:49:28.589 UTC Reload node ? [no,yes] yes result Admin VM graceful reload request on 0/RSP1 succeeded. sysadmin-vm:0_RSP1:eXR# RP/0/RSP1/CPU0:Mar 12 22:49:34.059 UTC: rmf_svr[402]: %PKT_INFRA-FM-3-FAULT_MAJOR : ALARM_MAJOR :RP-RED-LOST-ADMINNR :DECLARE :0/RSP1/CPU0: Confd is down RP/0/RSP1/CPU0:eXR#
Wait till the admin VM returns:
RP/0/RSP1/CPU0:eXR#0/RSP1/ADMIN0:Mar 12 22:59:30.220 UTC: envmon[3680]: %PKT_INFRA-FM-3-FAULT_MAJOR : ALARM_MAJOR :Power Module redundancy lost :DECLARE :0:
RP/0/RSP1/CPU0:Mar 12 22:59:33.708 UTC: rmf_svr[402]: %PKT_INFRA-FM-3-FAULT_MAJOR : ALARM_MAJOR :RP-RED-LOST-ADMINNR :CLEAR :0/RSP1/CPU0: