User`s guide

Chapter 16 Monitoring XgOS 309
■ You can reset statistics to zero to allow them to accumulate by using the optional
-clear argument.
■ Statistics are reset if the Fabric Board is reset, for example if the Oracle Fabric
Interconnect is power cycled.
▼ Determine the IB Path Continuity
As an option, if you have root access to the Oracle Fabric Interconnect, you can log
in as root and issue the ibclear errors first, then the ibcheckerrors command
to reset all the IB error counters. By doing so, you will have a fresh statistics
gathering session. This step is not mandatory, but if you do it, you will need to log
out and log back in to the Oracle Fabric Interconnect as admin to run the show
diagnostics ib-path command.
1. Note which server(s) are experiencing problems.
2. Log in to the Oracle Fabric Interconnect as admin user.
3. Issue the show diagnostics ib-path command without any qualifiers and
press Tab or ? to list all servers to verify that the server is in the resulting list.
In the following example, the server name is “brick.”.
show diagnostics ib-path
Possible completions:
<wildcard> Wildcarded name
IBM-1 Physical server running
Windows/6.1.7600/x64-2.7.0.269
IBM-2 Physical server running
Windows/6.1.7600/x64-2.7.0.269
MCKENZIE Physical server running
Windows/6.1.7600/x64-4.0.0.45
ROGUE Physical server running
Windows/6.0.6002/x64-2.7.0.47
WILLAMETTE Physical server running
Windows/6.1.7600/x64-2.7.0.269
arlo Physical server
brack Physical server running
VMware/ESX-4.0.0:xg-branch.3.8.0.esx4-/671
brick Physical server running
VMware/4.0.0:xg-2.6.5-2/x86_64
colusa Physical server running
Linux/2.6.18-53.el5:xg-2.6.7_RC2A/x86_64
daisy Physical server running
VMware/4.0.0:xg-2.6.5-2/x86_64
frack Physical server running
VMware/ESX-4.1.0:xg-3.1.0-r5722.ESX4.1/260247