SPARC T5-8
 

PLA-Support>Sun Systems>SPARC>CMT>SN-SPARC: T5
 

In this Document

Symptoms

Cause

Solution

References

Created from <SR 3-12548901311>

Applies to:

Symptoms

On servers that use ILOM Firmware on their Service Processors, the following Warning can be observed: /coredump is almost full

a. On the ILOM command line interface, the mentioned Warning can be seen after executing the command:

-> show /SP/logs/event/list

101 Tue Apr 19 11:13:02 2016 Fault Warning minor
Warning: /coredump is nearly full
100 Tue Apr 19 11:02:31 2016 Fault Warning minor
Warning: /coredump is nearly full

b. On the Host Operating System, the corresponding Warning message can be seen after executing the command
# grep '/coredump' /var/adm/messages
<timestamp> <hostname> SC Alert: [ID 876082 daemon.notice] Fault | minor: Warning: /coredump is almost full

Cause

The directory "/coredump" is used on the Service Processor's internal file system to store coredumps that are generated by processes
that run on the Service Processor's own Operating System. As soon as there is less than 1/8 of free space available in the "/coredump"
directory, the described Warning will be generated.

If firmware is < 3.2.5.x The root cause of dumps could be the bug : 20765548 FAILED to get target "/SYS/MB"

ILOM Dump files will show ilom/traces/ #####

@coredump@sp_trace@core@dump.ORACLESP-AK00186577.fetd#ip_xprt.core.23832.gz Tue Apr 19 09:38:22 2016
@coredump@sp_trace@core@dump.ORACLESP-AK00186577.fetd#ip_xprt.core.24490.gz Tue Apr 19 09:38:23 2016
@coredump@sp_trace@core@dump.ORACLESP-AK00186577.fetd#ip_xprt.core.24160.gz Tue Apr 19 09:38:22 2016

Solution

1) Reset the Service Processor:

-> reset /SP (this will not affect to Solaris)

2) Perform a firmware upgrade to last one (>=3.2.5.x) (downtime needed)

3) Reset the Service Processor:

-> reset /SP (this will not affect to Solaris)

References

This solution has no attachment