Sun Enterprise[TM] 10000 (E10K): domain does not come up after reboot: "Cannot read CBE information from snmpd."

Asset ID:	1-72-1004433.1
Update Date:	2017-12-06
Keywords:

Solution Type Problem Resolution Sure

Solution 1004433.1 : Sun Enterprise[TM] 10000 (E10K): domain does not come up after reboot: "Cannot read CBE information from snmpd."

Applies to:

Sun Enterprise 10000 Server - Version All Versions and later
All Platforms

Symptoms

Error on ssp: 'Cannot read CBE information from snmpd.'

Domain does not come up after reboot.

Dec 3 15:01:24 e10000 syslog: [ID 702911 local0.notice] bringup-(dom02): NOTICE: bringup.c, 2799: BRINGUP COMMAND STARTED for domain dom02
Dec 3 15:01:24 e10000 syslog: [ID 702911 local0.error] bringup-(dom02): ERR: ssp_snmp.c, 363: Snmp_Get(platformAmbientTemp.0) error; error status 2
Dec 3 15:01:24 e10000 syslog: [ID 702911 local0.error] bringup-(dom02): ERR: bringup.c, 2810: Cannot read CBE information from snmpd.
Dec 3 15:01:24 e10000 syslog: [ID 702911 local0.notice] bringup-(dom02): NOTICE: bringup.c, 2812: BRINGUP COMMAND FINISHED
Dec 3 15:03:39 e10000 syslog: [ID 702911 local0.error] bringup-(dom02): ERR: ssp_snmp.c, 363: Snmp_Get(confNumProc.0) error; error status 2
Dec 3 15:03:39 e10000 last message repeated 5 times
Dec 3 15:05:13 e10000 syslog: [ID 702911 local0.notice] bringup-(dom02): NOTICE: bringup.c, 2799: BRINGUP COMMAND STARTED for domain dom02
Dec 3 15:05:13 e10000 syslog: [ID 702911 local0.error] bringup-(dom02): ERR: ssp_snmp.c, 363: Snmp_Get(platformAmbientTemp.0) error; error status 2
Dec 3 15:05:13 e10000 syslog: [ID 702911 local0.error] bringup-(dom02): ERR: bringup.c, 2810: Cannot read CBE information from snmpd.
Dec 3 15:05:13 e10000 syslog: [ID 702911 local0.notice] bringup-(dom02): NOTICE: bringup.c, 2812: BRINGUP COMMAND FINISHED

Changes

Cause

The communication with the Control Board was not working.

Solution

Reboot the SSP (System Service Processor) or restart the SSP services, and wait 5-8 minutes to make sure all SSP daemons are started, then do bringup -A on the SSP for the domain.

Rebooting the SSP will not help if the CBE/CBS is out of memory and stuck, in this case a power cycle of the platform is needed.
We will have to verify the LED status in this case, please check the LEDs.

Pay particular attention to the cb_reset commands in /var/opt/SUNWssp/adm/messages as it is booting.
Problems with cb_reset will necessarily cause problems with CBE.

Troubleshoot cb_reset problems as you would an ARP/RARP issue.

When doing a CB reset we have to be careful by verifying where/which is the actual source that provides the hardware clock, if we are not careful will cause the platform to go down with a hardware clock failover event.

Check for the latest patches for the SSP on
"Sun SPARC Sun Enterprise 10000 Server (Starfire) Patches/Firmware" (Doc ID 1385502.1)

Attachments

This solution has no attachment