I'm setting up a UCS environment with a pair of MDS 9148's for the first time.  The fabric interconnects are 6120XP's with the 6 port 8Gb fc expansion modules.  Here is what I've done: I enabled npiv on both 9148's, I've assigned the FC ports on the fabric interconnects to the appropriate VSAN, I've assigned the ports on the 9148's to the appropriate VSAN, I've created a service profile template with vnic's and vhba's (the vhba's also being assigned to the appropriate vsan, and then I created two service profiles from the templates. I see that the two machines I created have logged into the fabric as I can see the wwn's in fabric manager.  I then created the alias and zones and saved the configuration.  I went to our storage managment software and allocated a LUN to each server.  Each server will be an ESXi server that is booting from SAN.  So I mount the ESXi iso to the machines and boot to the iso.  The machines boot to the iso and I can install ESXi onto the LUN (the machines see the boot LUN).  There is no local disk in these machines.  ESXi successfully installs and the machines reboot and I start configuring ESXi.  I go to lunch and come back and the ESXi servers have lost connectivity to the SAN.  I validate that ports on the 9148's are in the appropriate vsan, and the fabric interconnect ports are in the appropriate vsan, and the vhba template is using the appropriate vsan.  In Fabric Manager the connections from UCS to the individual servers I setup have a red X through them and when I click on the red X it takes me the connection of the vhba and says unavailableNPVUpstreamPort and the Port VSAN says 1, which is not the appropriate VSAN.  How did this get like this and how to I address it?  All the ports on the 9148's and fabric interconnects are in the appropriate vsan. ... View more

I setup a WWPN pool with the assignment order of default.  I would like to change it to sequential, but was wondering what sort of impact that would have to the environment that is already using that mac pool.  There are several machines already using some of the WWPN's and those I'll have to live with not using sequential until I can recong them.  It's not a change in UCS that would require me to reboot machines is it?           ... View more

I'm in the process of setting up UCS.  The default native vlan has a vlan ID of 1 in UCS.  Our native vlan is 1000.  So I setup a new vlan with the vlan ID of 1000 and set it as the natvie VLAN.  I cannot delete the VLAN default (1) even though it isn't the native vlan anymore because UCS won't let me.  We use VLAN id 1 for some of our corporate servers so I can't create a vlan with that ID without an overlap.  Since it's not being used as the native vlan anymore can I go ahead and use VLAN default (1) or is there some issue with me using that vlan? Additionally, one other question in regard to the natvie vlan.  I setup another UCS environment and have a few ESXi servers running on it with some active vm's.  When I setup UCS I added a vlan for our companines native vlan (vlan id 1000), but I forgot to set it as the Native VLAN.  So VLAN default (1) is still listed as the Native VLAN.  What implications would there be if I changed the Native VLAN to the vlan I setup (vlan id 1000) while there are running ESXi servers and virtual machines.  Neither the ESXi servers or vm's are using either on of those vlan's in service profiles and vnic templates. ... View more

We have a seperate physical network for our DMZ and a seperate physical network for some other isolated hosts.  Most of these machines in the two environments are virtual machines residing on 2U rack server machines running ESXi.  We are getting some UCS to replace these rack servers.  We'll have one set of fabric interconnects (6296's) and a couple chassis's with blades for ESXi.  My question is can I take uplink ehternet ports and connect them to the two different network environments.  For example, could I take ports 37 and 38 on both interconnects and connect them to the switches in our DMZ network and port 39 and 40 on both interconnects and connect them to swithces on the other network.  Would I be able to setup the ESXi server to connect to vlan's in both of the seperate networks as long as the trunks have the vlans on them? ... View more

I got some thermal alerts from a 2104XP IOM last week.  I investigated and saw no thermal issues in our datacenter, nothing was blocking airflow, and not other components were experiencing any type of thermal issue.  So I opened a case with Cisco TAC.  The looked through some tech support logs I sent them and indicated that the IOM's need to be replaced.  I asked if this was a disruptive process and they indicated - in theory it is not.  That answer didn't make me feel all that confident in replacing them.  Has anybody replaced both IO modules in a chassis before.  It is my understanding to just unplug the cables and pull out the IOM, insert the IOM, monitor it's progress in discovery and firmware update / downgrade, and once it's backup to proceed with the next one.  Is there anyway of validating that an IOM is appropriately functioning?  Or are there any specific steps I'm missing. It's impossible for me to take an outage to do this change.  The chassis hosts several hundred production virtual machines.              ... View more

We are starting to grow our UCS environment.  Due to that we are using more and more IP's in the pool we set aside for Management IP addresses.  Presently, our fabric interconnects reside in the same subnet (class C) as the pool we are using for Management IP's.  It looks as though each blade uses two IP's out of the Management Pool so we are exhausting the IP's that we set aside pretty quickly.  When I went to setup the UCS environment initially I went to use a subnet for the Management Pool that was different than the subnet being used for the fabric interconnects.  Unfortuantely, I was unable to connect to the KVM functionality with the Management Pool in a different subnet that the fabric interconnects.  When I changed the Management Pool IP's to the same subnet as the fabric interconnects the issue went away.  So I take it that the KVM functionality comes through the management interfaces of the fabric interconnects. My question is how do I provide a different subnet for my Management Pool.  I will run out of IP's in this one subnet eventually and will have to add another one in that is different from what the fabric interconnects are using.  Do I have to set the ports that the fabric interconnect management interfaces are plugged into for vlan tagging and then change the network configuration of the fabric interconnects? ... View more

I'm not a SAN guy nor an expert in UCS.  I've inherited our SAN environment due to our Storage Engineer leaving.  We started out with UCS in the 1.4 version that didn't have fiber channel port channels I believe.  We are on 2.1(1d) now.  Right now we have individual fc uplinks plugged into brocade silkworm switches attached to HDS mid tier arrays.  What advantages are there to creating fiber channel port channels?  Do I need to do anything on the brocade switches to create a port channel.  When our network people create port channels on the network switches that have to set those ports up specifically.        ... View more

I will be replacing a 6120 with a 6296 in the near future.  I have the Upgrading Cisco UCS B-Series Hardware pdf which gives an overview of the process for replacing a 6120 with a 6248.  The 6120 we have has the 6 port 8Gb fiber channel expansion module.  I was wondering if anybody has had experience doing this replacement and if there are any gotchas or advice that you can offer in this replacement.       ... View more

We have a couple 6120XP fabric interconnects setup in a cluser.  The expansion modules have the 6 port 8Gb fiber channel expansion modules in them.  Presently, we are only using two of the 8Gb ports per fabric interconnect.  They are connected to ports on a Brocade Silkwork 48k.  These 6120XP's have been in production for two years now so we've been accessing storage through our existing configuration for several years now.  Our storage administrator who set this up is no longer with us. When I go into Zone Admin on the swithes I do not see the WWN's for the SFP's that are inserted into teh 8Gb ports on the expansion modules?  Should I see them.  I have a vague recollection that with NPIV I may not be seeing those.  When I spin up a new blade I will see the WWN's for the blade.  What I'm looking to do is add a couple more active ports to increase the bandwidth.  I went to check out the Zones on the switches and I can't find any configuration for UCS on the switches other than the individual blades.  Do I just need to plug additional ports into my SAN switches and then enable the ports in UCS to be good to go?                ... View more