07-17-2023 09:55 PM
Is there a PowerShell script or a GUI in UCSM which I can use to check which servers in my UCSM environment do not have redundant connections to the FIs?
For example, I have one server (Cisco UCS C220 M4S) with only 1 CPU (Intel(R) Xeon(R) E5-2630 v3) and thus only supports one adapter. Someone installed 2x Cisco UCS VIC 1225 adapters in the server however only one is usable due to the one CPU. I recently found, by accident, that the server only had one viable connection to the FEXs and I was able to fix the issue by replacing the SFP, however I would like a way to check all servers to ensure they all have redundant network connections.
Thanks.
Solved! Go to Solution.
07-18-2023 05:47 AM
Under each physical server, the VIF paths display in the UI will show you which vnics/vhbas are connected to FI A or FI B and their link state
Or it can be obtained by SSHing into the UCS CLI with "show service-profile circuit" or with powershell:
Get-UcsServer | Get-UcsFabricLocale | Get-UcsFabricPath | Get-UcsDcxVc | Select dn,switchid,operstate,adminstate,vnic,linkstate | Where { $_.vnic } | Sort dn | ft
In this powershell example output chassis 1 blade 1 only has connections via FI B due to an issue with FI A uplinks being unavailable:
Dn SwitchId OperState AdminState Vnic LinkState
-- -------- --------- ---------- ---- ---------
sys/chassis-1/blade-1/fabric-A/path-1/vc-2326 A link-down enabled eth2 down
sys/chassis-1/blade-1/fabric-A/path-1/vc-2327 A link-down enabled eth3 down
sys/chassis-1/blade-1/fabric-A/path-1/vc-2329 A link-down enabled vhba_A down
sys/chassis-1/blade-1/fabric-B/path-1/vc-2325 B active enabled eth1 up
sys/chassis-1/blade-1/fabric-B/path-1/vc-2328 B active enabled eth0 up
sys/chassis-1/blade-1/fabric-B/path-1/vc-2330 B active enabled vhba_B up
Also would recommend checking inside the operating system as well for redundancy, sometimes there might only be 1 nic configured on the vswitch at the os level or the wrong vnics were paired together in a team at the os level etc..
07-18-2023 05:47 AM
Under each physical server, the VIF paths display in the UI will show you which vnics/vhbas are connected to FI A or FI B and their link state
Or it can be obtained by SSHing into the UCS CLI with "show service-profile circuit" or with powershell:
Get-UcsServer | Get-UcsFabricLocale | Get-UcsFabricPath | Get-UcsDcxVc | Select dn,switchid,operstate,adminstate,vnic,linkstate | Where { $_.vnic } | Sort dn | ft
In this powershell example output chassis 1 blade 1 only has connections via FI B due to an issue with FI A uplinks being unavailable:
Dn SwitchId OperState AdminState Vnic LinkState
-- -------- --------- ---------- ---- ---------
sys/chassis-1/blade-1/fabric-A/path-1/vc-2326 A link-down enabled eth2 down
sys/chassis-1/blade-1/fabric-A/path-1/vc-2327 A link-down enabled eth3 down
sys/chassis-1/blade-1/fabric-A/path-1/vc-2329 A link-down enabled vhba_A down
sys/chassis-1/blade-1/fabric-B/path-1/vc-2325 B active enabled eth1 up
sys/chassis-1/blade-1/fabric-B/path-1/vc-2328 B active enabled eth0 up
sys/chassis-1/blade-1/fabric-B/path-1/vc-2330 B active enabled vhba_B up
Also would recommend checking inside the operating system as well for redundancy, sometimes there might only be 1 nic configured on the vswitch at the os level or the wrong vnics were paired together in a team at the os level etc..
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide