Sun Datacenter InfiniBand Switch 36
 
Exadata Database Machine X2-2 Full Rack
 

PLA-Support>Sun Systems>SAND>Network>SN-SND: Sun Network Infiniband
 

In this Document

Symptoms

Cause

Solution

Applies to:

Symptoms

 Running ibswitchcheck.sh post on a leaf switch falsley reports the switch as a spine

[root@ib2 tmp]# ./ibswitchcheck.sh post

 Current version of switch is: 1.3.3-2 and is a spine            <<<<-----------------------
 Switch target version is: 1.3.3-2
 [SUCCESS] Switch is at target patching level

 [SUCCESS] Hostname ora2sw-ib2.central.hq.internal passed validation
 in /etc/hosts and /etc/sysconfig/network-scripts/ifcfg-eth0

 [SUCCESS] Service ntpd is running
 [ERROR] Service opensmd is not running
 Action: Start opensm service by issuing 'enablesm'

 [SUCCESS] Number of ntp servers 2 meets minimal requirement of 1
 ntp server: 12.12.1.0
 ntp server: 10.25.21.232

 [ERROR] opensm.conf failed all or some of the validation
 [SUCCESS] routing_engine value in /etc/opensm/opensm.conf is ftree
 [SUCCESS] controlled_handover value in /etc/opensm/opensm.conf is TRUE
 [SUCCESS] sminfo_polling_timeout value in /etc/opensm/opensm.conf is
 1000
 [SUCCESS] polling_retry_number value in /etc/opensm/opensm.conf is 5
 [ERROR] sm_priority must equal 8, value in /etc/opensm/opensm.conf is 5
 Action: issue disablesm, set priority using 'setsmpriority 8', issue
 enablesm

 Manually validate the following entries
 Date:(MM/DD/YY) 01/16/13
 Time:(HH:MM:SS) 13:10:29
 Timezone: EST

 [SUCCESS] /conf/configvalid is 1

 Overall status of post check is ERROR

Cause

The script wrongly reports that this is a spine switch if there are NO hosts/nodes on the fabric.

If the ib fabric is not up or the cables are not connected you may see the ERROR above.

Solution

Ignore the error message if the fabric is not up or if the nodes/hosts are not UP on the IB fabric.

Insure that at least some of the nodes are up on the network before running the script.

This solution has no attachment