UCS - san-port-channel over vpc

Hi,
I'm hoping someone can help out with this,
I have a setup with 2 Nexus 5548's connected over a vpc peer link and a keepalive, These two nexus switch's have a standard VPC connection to an upstream 6500. this is working fine,
Then there are multiple native FC connections from the 5548's to 6248's. I know there is the option to configure FC port-channel's on the UCS 2.0 but is it possible to put fc interfaces of the nexus 5548 into a san port-channel across a vpc between the the nexus 5548's.
so for example, can i port-channel the following across a vpc or is a san port-channel restricted to ports on the same nexus switch only.
Fabric A fc1/31 & fc1/32 uplinked to -> NX5K-1 Port fc1/31 & NX5K-2 Port fc1/32 (san port-channel 10 over a VPC)
Fabric B fc1/31 & fc1/32 uplinked to -> NX5K-1 Port fc1/32 & NXFK-2 Port fc1/31 (san port-channel 11 over a VPC)
Thanks,
Ray.

This is not possible. Vpc is layer 2 only ethernet port channel. San port channel cannot be built accross vpc
Sent from Cisco Technical Support iPhone App

Similar Messages

UCS FI 6248 to Nexus 5548 San port-channel - not working

Hi all,
I'm sure I am missing something fairly obvious and stupid but I need several sets of eyes and help.
Here is the scenario:
I want to be able to create san port-channels between the FI and Nexus. I don't need to trunk yet as I can't even get the channel to come up.
UCS FI 6248:
Interfaces fc1/31-32
Nexus 5548
interfaces fc2/15-16
FI is in end-host mode and Nexus is running NPIV mode with fport-channel-trunk feature enabled.
I'm going to output the relevants configurations below.
Nexus 5548:
NX5KA(config)# show feature | include enabled
fcoe                  1         enabled
fex                   1         enabled
fport-channel-trunk   1         enabled
hsrp_engine           1         enabled
interface-vlan        1         enabled
lacp                  1         enabled
lldp                  1         enabled
npiv                  1         enabled
sshServer             1         enabled
vpc                   1         enabled
interface san-port-channel 133
channel mode active
no switchport trunk allowed vsan all
switchport trunk mode off
interface fc2/15
switchport trunk mode off
channel-group 133 force
no shutdown
interface fc2/16
switchport trunk mode off
channel-group 133 force
no shutdown
NX5KA# show vsan membership
vsan 1 interfaces:
    fc2/13            fc2/14
vsan 133 interfaces:
    fc2/15            fc2/16            san-port-channel 133
vsan 4079(evfp_isolated_vsan) interfaces:
vsan 4094(isolated_vsan) interfaces:
NX5KA# show san-port-channel summary
U-Up D-Down B-Hot-standby S-Suspended I-Individual link
summary header
Group Port-      Type    Protocol               Member Ports
       Channel
133    San-po133 FC      PCP      (D) FC       fc2/15(D)    fc2/16(D)
UCS Fabric Interconnect outputs:
UCS-FI-A-A(nxos)# show san-port-channel summary
U-Up D-Down B-Hot-standby S-Suspended I-Individual link
summary header
Group Port-      Type    Protocol               Member Ports
       Channel
133    San-po133 FC      PCP      (D) FC       fc1/31(D)    fc1/32(D)
UCS-FI-A-A(nxos)#
UCS-FI-A-A(nxos)# show run int fc1/31-32
!Command: show running-config interface fc1/31-32
!Time: Fri Dec 20 22:58:51 2013
version 5.2(3)N2(2.21b)
interface fc1/31
switchport mode NP
channel-group 133 force
no shutdown
interface fc1/32
switchport mode NP
channel-group 133 force
no shutdown
UCS-FI-A-A(nxos)#
UCS-FI-A-A(nxos)# show run int san-port-channel 133
!Command: show running-config interface san-port-channel 133
!Time: Fri Dec 20 22:59:09 2013
version 5.2(3)N2(2.21b)
interface san-port-channel 133
channel mode active
switchport mode NP

!Command: show running-config interface san-port-channel 133
!Time: Sat May 16 04:59:07 2009
version 5.1(3)N1(1)
interface san-port-channel 133
channel mode active
switchport mode F
switchport trunk mode off
Changed it as you suggested...
Followed the order of operations for "no shut"
Nexus FC -> Nexus SAN-PC -> FI FC -> FI SAN-PC.
Didn't work:
NX5KA(config-if)# show san-port-channel summary
U-Up D-Down B-Hot-standby S-Suspended I-Individual link
summary header
Group Port-      Type    Protocol               Member Ports
       Channel
133    San-po133 FC      PCP      (D) FC       fc2/15(D)    fc2/16(D)
NX5KA(config-if)#
Here is the output as you requested:
NX5KA(config-if)# show int san-port-channel 133
san-port-channel 133 is down (No operational members)
    Hardware is Fibre Channel
    Port WWN is 24:85:00:2a:6a:5a:81:00
    Admin port mode is F, trunk mode is off
    snmp link state traps are enabled
    Port vsan is 133
    1 minute input rate 1256 bits/sec, 157 bytes/sec, 0 frames/sec
    1 minute output rate 248 bits/sec, 31 bytes/sec, 0 frames/sec
      3966 frames input, 615568 bytes
        0 discards, 0 errors
        0 CRC, 0 unknown class
        0 too long, 0 too short
      2956 frames output, 143624 bytes
        0 discards, 0 errors
      46 input OLS, 41 LRR, 73 NOS, 0 loop inits
      257 output OLS, 189 LRR, 219 NOS, 0 loop inits
    last clearing of "show interface" counters never
    Member[1] : fc2/15
    Member[2] : fc2/16
NX5KA(config-if)#
NX5KA(config-if)# show int brief
Interface Vsan   Admin Admin   Status          SFP    Oper Oper   Port
                  Mode   Trunk                          Mode Speed Channel
                         Mode                                 (Gbps)
fc2/13     1      auto   on      sfpAbsent        --     --           --
fc2/14     1      auto   on      sfpAbsent        --     --           --
fc2/15     133    F      off     init             swl    --           133
fc2/16     133    F      off     init             swl    --           133

Creating san-port-channel on 6248 using CLI

I can create the san port channel using UCS manager. But I would like to know the syntax using CLI. So far I was able to figure out the following. One thing missing was moving the port channel from the default VSAN(1) to a different VSAN i.e VSAN 10 in my case. I have looked at the CLI guide, but have not found the commands syntax. Appreciate if anyone can post the syntax
### Create Port Channel ###
scope fc-uplink
scope fabric a
create port-channel 1
enable
set name port-channel-1
set adminspeed auto
commit-buffer
end
### Add interfaces to Port Channel ###
scope fc-uplink
scope fabric a
scope port-channel 1
create member-port 1 29
exit
create member-port 1 30
commit-buffer
end

Thanx Wdey. That's exactly what I was looking for. Here is the syntax which works for me.
scope fc-uplink
scope fabric a
create port-channel 1
enable
set name port-channel-1
set adminspeed auto
commit-buffer
end
scope fc-uplink
scope fabric a
enter port-channel 1
enable
enter member-port 1 29
enable
exit
enter member-port 1 30
enable
exit
set adminspeed auto
exit
enter vsan VSAN10 10 10
set fc-zoning disabled
set fcoe-vlan 10
set id 10
enter member-port-channel a 1
exit
localize
exit
end
commit-buffer

SAN Port Channel between 6120 and Brocade

Has anyone setup or know how to setup a SAN Port Channel between 6120s and a Brocade FC switch? I know that it requires UCS firmware 1.4 but I don't know what the Brocade requirements are.

The SAN port channel is Cisco proprietory. You cannot setup SAN port channel with Brocade. Sorry

SAN Port-Channel between Nexus 5000 and Brocade 5100

I have a Nexus 5000 running in NPV mode connected to a Brocade 5100 FC switch using two FC ports on a native FC module in the Nexus 5000. I would like to configure these two physical links as one logical link using a SAN Port-Channel/ISL-Trunk. An ISL trunking license is already installed on the Brocade 5100. The Nexus 5000 is running NX-OS 4.2(1), the Brocade 5100 Fabric OS 6.20. Does anybody know if this is a supported configuration? If so, how can this be configured on the Nexus 5000 and the Brocade 5100? Thank you in advance for any comments.
Best regards,
Florian

I tried that and I could see the status light on the ports come on but it still showed not connected.
I configured another switch (a 3560) with the same config and the same layout with the fiber and I got the connection up on it. I just cant seem to get it on the 4506, would it be something with the supervisor? Could it be wanting to use the 10gb port instead of the 1gb ports?

MDS Code Level for UCS FC Port Channel/Trunking

What is the required MDS code level needed to support FC Port-Channeling/Trunking on a 9222i, 9148, or 9509/9513 w/ Gen2 line cards to UCS 1.4(3)?
I know the fport-channel-trunk feature is needed and was released in MDS 4.2 code but I have seen documents that this level of code wont support these features when used to connect to Cisco UCS.
Can someone confirm? Is NXOS 5.0(4) or higher required?

Doug,
I wrote the doc almost a year ago and it doesn't have the latest interop matrix support information. The current MDS NX-OS support is located in this doc: http://www.cisco.com/en/US/docs/switches/datacenter/mds9000/interoperability/matrix/intmatrx.html
While, I wish this was also duplicated in UCS docs, they are more or less taking the stance that 3rd party (Cisco MDS is this case) interop is maintained by the 3rd party. EMC, Netapp, or even Cisco.
It would be a nightmare updating UCS docs with 3rd party qualifications and support.
The short answer to your question is:
UCS 1.4(1), 1.4(2), 1.4(3) Is supported by MDS9000: 3.3(4a), 3.3(5), 4.2(1a), 4.2(1b), 4.2(3a), 4.2(5), 5.0(1a), 5.0(4), 5.0(4c)
NPV was introduced in MDS 3.3. f port-channel/trunk was introduced in MDS 4.2.
As Ram mentioned above, 4.2(5) is what I would recommend if I wanted to implement f port-channel and/or f port-trunk. I have also had recent experience with 4.2(5) and it worked great with a large UCS f port-channel/trunk implementation
The specific switch limitations I mentioned in my doc still exist. Only certain models of MDS will support f port-trunk. Almost all models will support f port-channel.
Dave

Difference between Port Channel and VPc

Hi Friends,
Could you please provide the difference between Port Channel and VPC.
Regards,
Zaheer

Read :)
http://www.cisco.com/c/en/us/products/collateral/switches/nexus-3000-series-switches/white_paper_c11-685753.html
Virtual PortChannel Technology
Virtual PortChannels (vPCs) allow links that are physically connected to two different Cisco® switches to appear to a third downstream device to be coming from a single device and as part of a single PortChannel. The third device can be a switch, a server, or any other networking device that supports IEEE 802.3ad PortChannels.
Cisco NX-OS Software vPCs and Cisco Catalyst® Virtual Switching Systems (VSS) are similar technologies. For Cisco EtherChannel technology, the term “multichassis EtherChannel” (MCEC) refers to either technology interchangeably.
vPC allows the creation of Layer 2 PortChannels that span two switches. At the time of this writing, vPC is implemented on the Cisco Nexus® 7000 and 5000 Series platforms (with or without Cisco Nexus 2000 Series Fabric Extenders).

Xconnect 20Gb Port-channel over MPLS Possible?

Can anybody confirm if you can establish an Xconnect between two 20Gb port-channels (2x 10Gig bundle). The core is 40Gig and running MPLS. We have many gigabit pseudowires running over it already.
Our intention is to terminate 10Gig server traffic on 4900M's and then have a 20Gb Port channel trunk link into a 6500 (which is part of the MPLS cloud). We would then issue the xconnect directly on the 20Gig port channel in order to traverse the MPLS Core.
Has anyone configured this before? Are there any caveats to watch out for? What sort of performance hit (if any does) xconnect functionality have on 10Gig Traffic.

Hi,
Yes it's supported but test it first (as always ;-) )
HTH
Laurent.

Port Channel over L2 MPLS links

Hello.
I was hoping that someone could over some suggestions on best practice or recommendations for configuring port channel to bundle to layer 2 MPLS links that we have.
We have racks in two geographically separated data centres, each rack has a stack of Cisco 3850 switches, and there are two 100Mbps layer 2 connections linking these stacks together. The links are provided by our supplier and runs over their core network (I can get more info on their setup if need be).
I had initially just configured a simple port channel bundling the two ports on "Switch A", and the same on "Switch B". However I have since noticed that one of the ports was put into error disabled state, and I have my doubts that we ever had 200Mbps throughput over the portchannel.
The logs showed:
UDLD-4-UDLD_PORT_DISABLED: UDLD disabled interface Gix/x/x, aggressive mode failure detected
%PM-4-ERR_DISABLE: udld error detected on Gix/x/x, putting Gix/x/x in err-disable state
How I understand is that one of ports recieved an unexpected (or incorrect) BPDU packet, and shut the port down as part of spanning tree process to stop a loop forming.
For the time being I have removed the port channel config and re-opened all the ports, so I believe one of the ports is now in blocking state.
Whilst redundancy for these links is great, ideally I would like to have the links bundled so I also get the benefit of higher throughput.
Any thoughts are greatly appreciated.
Thank you

Hi,
Thank you for your reply and suggestions. I have been reading up on layer 2 protocol tunnelling and I'm not sure if this will work for us. According to these guidelines for the catalyst 3550 (http://www.cisco.com/c/en/us/td/docs/switches/lan/catalyst3550/software/release/12-1_19_ea1/configuration/guide/3550scg/swtunnel.html#wp1006657) it states:
•If you enable PAgP or LACP tunneling, we recommend that you also enable UDLD on the interface for faster link-failure detection.
•Tunneling is not supported on trunk ports. If you enter the l2protocol-tunnel interface configuration command on a trunk port, the command is accepted, but Layer 2 tunneling does not take affect unless you change the port to a tunnel port or an access port.
•EtherChannel port groups are compatible with tunnel ports when the 802.1Q configuration is consistent within an EtherChannel port group.
(maybe its a different and is supported on 3850's)
Also, I cant use layer 3 over this link as I am looking to extend VLANs over the link so devices and virtual machines in each of the datacentre are in the same subnet and broadcast domain.
I will speak with the supplier to see what they are willing to do.
Thanks again.

Does Solaris 10 support cisco Virtual Port Channels over IP?

Does anyone know if Solaris 10 support cisco Virtual Port Channels open IP?

Hi user11114413,
The issue you are seeing actually has little to do with VIP, and more to do with there being multiple IP addresses for us to choose from on your box. For such multi-IP boxes, you'll want to tell us the IP to use, and in your case you want to tell as a VIP. This can be done either by editing your operational configuration file, and including an <address> element within the <unicast-listener> element, or via the tangosol.coherence.localhost system property. For example:
<unicast-listener>
    <well-known-addresses>
        <socket-address id="1">
            <address>1.2.3.260</address> 
            <port>8088</port>
        </socket-address>
    </well-known-addresses>
    <address>1.2.3.260</address> 
    <port>8088</port>
</unicast-listener>or
java ... -Dtangosol.coherence.localhost=1.2.3.260If you are using the same operational configuration on all nodes in your cluster then the system property approach is likely preferable, and would only be necessary on the two machines sharing the VIP.
As for using VIP or an extended WKA list, the choice is yours, either will work. If you do go the VIP route, it would obviously be a very bad idea to simultaneously use the same VIP and port at the same time from the two machines.
thanks,
Mark
Oracle Coherence

Create FI6248 SAN Port-Channel - No Operational Members

My problem is similar to this discussion:
https://supportforums.cisco.com/thread/2169563
The environment is UCS 2.1(1a), 6248s, 5548s.
The configuration is exactly the same as well as the symptoms. I already enabled fport-channel-trunk feature, and verified it with "show feature" command.
Really appreciate if someone can give some diretions to figure this thing out.
Thanks for your help.

Hello ,
What is the version on N5K ?
Does it work if you configure it as individual links ?
Can you please check if you are running OUI issue ?
http://tools.cisco.com/Support/BugToolKit/search/getBugDetails.do?method=fetchBugDetails&bugId=CSCty04686
HTH
Padma

MDS configuration for port channel

I've tried to set up the new FC port channel upload to a pair of MDS 9124s, but as I don't know enough about the MDS side I can't get the link to come up. Are there any references available anywhere that tell you blow by blow exactly how to configure the MDS side of the port channel? Or perhaps a sample working config?
Thanks

Simon,
A few items:
1) You are on 1.4.1i Balboa code
2) FI's are in FC switch mode as MDS 9124's don't support F-port channel (NPV/NPIV)
3) It's a lot easier the first time if you have matched speed sfp+ all around. Although, with the correct configs, unmatched speeds can be made to work.
If the FIs are in switch mode and your MDS is running a minimum 3.3 here are some configs that may help:
MDS side interfaces 4/11 and 4/12 are plugged into the UCS fc ports. First setup and ensure ISLs are working properly between UCS and MDS, then configure the port-channels
interface fc4/11
switchport rate-mode dedicated
switchport mode E
switchport trunk mode auto
no shutdown
interface fc4/12
switchport rate-mode dedicated
switchport mode E
switchport trunk mode auto
no shutdown
The UCS GUI Equipment -> Fabric Interconnects -> FI-A -> Set FC Switching Mode. This will cause both FI's to reboot into FC switch mode. This is VERY DISRUPTIVE, both FI's will reboot.
On the UCS CLI you should see this on the UCS fc ports connected to the MDS
cae-sj-ca3-A(nxos)# show running-config interface fc 2/1-2
!Command: show running-config interface fc2/1-2
!Time: Wed Oct 20 16:49:39 2010
version 4.2(1)N1(1.4)
interface fc2/1
switchport mode E
no shutdown
interface fc2/2
switchport mode E
no shutdown
Until you have VSAN trunks enabled, make sure the ports on the UCS and MDS are in the same VSAN. VSAN 1 for example.
Once you have working ISLs, then proceed to turn them into port-channels and enable VSAN trunking. Start with the MDS
Create a channel group, it should look like this
cae-sj-9506-1# show run interface port-channel 3
!Command: show running-config interface port-channel 3
!Time: Wed Oct 20 17:06:01 2010
version 5.0(1a)
interface port-channel 3
channel mode active
switchport mode E
switchport rate-mode dedicated
switchport trunk mode auto <-- VSAN trunking
Enable VSAN trunking on the MDS ISL interfaces, it'll look like this:
cae-sj-9506-1# show run interface fc 4/11-12
!Command: show running-config interface fc4/11-12
!Time: Wed Oct 20 17:07:05 2010
version 5.0(1a)
interface fc4/11
switchport rate-mode dedicated
switchport mode E
switchport trunk mode auto
no shutdown
interface fc4/12
switchport rate-mode dedicated
switchport mode E
switchport trunk mode auto
no shutdown
Add the channel group information to the MDS ISL interfaces:
Enter configuration commands, one per line. End with CNTL/Z.
cae-sj-9506-1(config)# interface fc 4/11-12
cae-sj-9506-1(config-if)# channel-group 3 force
fc4/11 fc4/12 added to port-channel 3 and disabled
please do the same operation on the switch at the other end of the port-channel,
then do "no shutdown" at both ends to bring it up
cae-sj-9506-1(config-if)# show run interface fc 4/11-12
!Command: show running-config interface fc4/11-12
!Time: Wed Oct 20 17:07:39 2010
version 5.0(1a)
interface fc4/11
switchport rate-mode dedicated
switchport mode E
switchport trunk mode auto
channel-group 3 force
no shutdown
interface fc4/12
switchport rate-mode dedicated
switchport mode E
switchport trunk mode auto
channel-group 3 force
no shutdown
Create the SAN port channel on the UCS side
SAN -> SAN Cloud -> Fabric A -> FC Port Channels -> Create Port Channel
After you create the SAN port channel in UCS, make sure and enable it. I also bounce the MDS port channel at this point.
If all is correct, in a few agonizing minutes, the port-channel will be formed and passing data.
cae-sj-ca3-A(nxos)# show interface san-port-channel 1
san-port-channel 1 is trunking
    Hardware is Fibre Channel
    Port WWN is 24:01:00:0d:ec:d3:5d:c0
    Admin port mode is E, trunk mode is on
    snmp link state traps are enabled
    Port mode is TE
    Port vsan is 1
    Speed is 8 Gbps
    Trunk vsans (admin allowed and active) (1,10,26,50,66,100-101,103,123,222,24
0)
    Trunk vsans (up)                       (1,10,50,100,103)
    Trunk vsans (isolated)                 (26,66,101,123,222,240)
    Trunk vsans (initializing)             ()
    5 minute input rate 2312 bits/sec, 289 bytes/sec, 2 frames/sec
    5 minute output rate 1440 bits/sec, 180 bytes/sec, 2 frames/sec
      2669 frames input, 194760 bytes
        0 discards, 0 errors
        0 CRC, 0 unknown class
        0 too long, 0 too short
      2677 frames output, 158316 bytes
        0 discards, 0 errors
    0 input OLS, 1 LRR, 0 NOS, 0 loop inits
Again, make sure you have the ISLs up an running first before configuring the port-channels. It makes troubleshooting much easier.
Let me know if you need any help.

PIN GROUP, Port Channel

Hi, Cisco Gurus:
Need your authoritative take on the following: Ethernet and FC is all End-Host Mode.
1. If client is using FC Port Channel on all of the FC Uplinks for the 2 Fabric Interconnects with a SINGLE VSAN on each FI, MUST THEY USE SAN PIN GROUP? Currently, they are pinning the vHBA template to this SAN Pin Group. vHBA-Fabric A pins to FC-PIN-GROUP-A and vice-versa,
2.
If FC Port Channel is being used as in 1, from WHICH UCSM option should the VSAN be created?
Should VSAN be STILL created under the Individual Fabric A and B as before?
Or
It have to be Global VSAN or Dual VSAN outside the Fabric A or Fabric B?
3.If client is using Ethernet Port Channel for each Fabric Interconnet or each fabric, must they use LAN PIN GROUP?
Currently, they are pinning the vNIC template to this Pin Group (Fabric A & B). vNICA-Fabric A pins to LAN-PIN-GROUP-A and vice-versa.
Something which I have SELDOM SEE before.
As usual, I know it is easy meat for Cisco gurus. Thanks, PAL.
So Long.
SiM

All VSANs in the screenshot above are Global. Notice they're not created under the Fabric A/ B section.
Unlike "VLANs", which are almost always Global and normally accessible from each Fabric Interconnect, VSANs are different. Best practice for storage has keeping each SAN Fabric separate. Keeping this in mind, does it make more sense to create VSANs as global or under their respective Fabric interconnects? (Under the separate Fabric Inteconnects of course. The only use case for using Global VSANs is as Craig pointed out - when there is a single Storage switch northbound that both Fabric Interconnects uplink to.
Where this gets confusing is when you're assigning vHBA's to a VSAN.
Ex.
Let's say, as your customer has done, they've created their two VSAN 10 and 20 as "Global". For comparison sake I've also created VSAN 100 just under Fabric-A and VSAN 200 just under Fabric-B. Their VSANs would looks like:
Now, when creating an vHBA under your service profile, you need to assign it to a VSAN. Take a look:
So depending which Fabric (A/B) they want the vHBA connected to the options will change. You can see that by creating the VSANs 10 and 20 as "Global" they appear when either Fabric is selected - Conversly you can see the two 100 and 200 I created under each Fabric, only appear when that fabric is selected.
So what? Well if your upstream SAN switches only use one of the respective VSANs configured and only connect to one Fabric Interconnect, then you can potentially "blackhole" your storage traffic by assigning it to the wrong Fabric Interconnect. Creating the VSANs only under the appropriate Fabric Interconnect prevents this possibility.
SAN Port-Channels have no bearing on this. All a SAN Port Channel does is make mulitple FC Uplinks appear as one - which requires SAN Port Channels configured upstream also. Leaving the FC Uplinks as individual just lets UCS dynamically pin virtual interfaces across the FC uplinks in a round-robin fashion.
In your case you have no need for Pin-Groups either. As Jeremy pointed out, unless you have a specific need for them, it's safer to not use them.
Regards,
Robert

MDS9513 Add ISL to Port Channel problem.

Hi Experts,
I have a problem when add a new ISL to an existing Port Channel config.
The ISL port is up state and correctly configured like another ports in port channel.
Looks bellow the error:
MDS1300M6A# show port-channel summary
Interface                 Total Ports        Oper Ports        First Oper Port
port-channel 1                 4                 4                  fc12/48
port-channel 2                 2                 2                  fc6/47
MDS1300M6A# show port-channel database
port-channel 1
    Administrative channel mode is active
    Operational channel mode is active
    Last membership update failed: port not compatible [Resources Unavailable]
    First operational port is fc12/48
    4 ports in total, 4 ports up
    Ports:   fc12/48 [up] *
             fc10/48 [up]
             fc11/48 [up]
             fc9/48   [up]
Anyone seen this?
Tks
Wellington

That did the trick - specifically setting the port channel to rate-mode shared. Certainly solved the problem as far as getting all the deisred ports up in the channel, but where I still have a knowledge gap is the operational difference between dedicated and shared. Are there any Cisco docs (beyond the command reference guide to toggle the feature) that explains why you'd choose one versus the other, and what it means from a design perspective?
On a different note, while I'm no expert, here's some background to consider for others who may encounter the same issue (whether on UCS or a Nexus 5K). I (and our network team) come from an IOS background, so while the IOS "parts" of NX-OS are pretty familiar, the SAN "parts" of NX-OS are a little new. We're also more CLI oriented, and while not impossible to configure via CLI, in our environment, the GUI for UCS seems to be more popular across the board - and we're still learning that piece as well.
Our storage team is very familiar with SAN-OS, so they're very comfortable with the SAN "parts" of NX-OS. But, they're very used to configuring the MDS switches via the GUI, which is what got us off track a bit. In our scenario, creating a SAN port channel on the MDS via the GUI didn't give us the option for setting "switchport rate-mode shared" Not saying it's not there in the GUI, but we certainly couldn't find it.
Long story short, if the configuration of the SAN port channel on the MDS is in question, check it via the CLI and make any necessary changes there as a quick workaroud.

FC Port Channel, Storage Cloud???

Hi. Cisco gurus:
May I have the authoritative answers on the following?
1. What is the Use Case of FC Port Channel under SAN Cloud assuming I have the typical Cisco VSAN setup-1 VSAN tied to 1 SAN Switch and 1 FCoE VLAN and another VSAN with diff. VSAN and FCoE ids, etc?
1.1 What is the commands needed in MDS switch to match the FC Port Channel in UCS so as to make it works, so to speak.
1.2 Will it make any differences between the End-Host Mode and Switch Mode?
2.
What is the benefits of FC Port Channels? Is it REALLY necessary at all assuming the same as 1 above with 2 Uplink FC Per Fabric Interconnect?
3.
What is the Use Case of Storage Cloud? Is it ONLY for direct-attached FC storage to our Fabric Interconnect?
Thank again. Help me to help Cisco to sell more UCS.
SiM

Sim, assuming you aren't using pin groups UCS will automatically re-pin the server vNIC/vHBA in the evnet of a LAN/SAN uplink failure. This re-pinning will take a few seconds but that isn't enough that you would notice the down time.
With a port channel no re-pinning has to happen as there is only a single logical uplink. When physical connection in a port channels fails there is a sub-second convergence time.
Port channels offer better band-width utilization because all servers are pinned to a single logical uplink.
If you do not have MDS FC switches then you can't create a SAN port channel. For non-MDS environments with 2 or more FC uplinks UCS will perform dynamic pinning of server vHBAs to those uplinks in a round-robin fashion. If one of the uplinks fails UCS will re-pin those vHBAs to another uplink.
Most of our UCS customers don't need VSAN trunking as they ony have 1 VSAN northbound from each fabric but if you did have more the one VSAN on your fabric then you could provision a vHBA to be in that VSAN. For example lets say you had 2 storay arrays on each SAN fabric and each was in its own VSAN. You could create vHBAs in the different VSANs so that you could build a server that has access to LUNs in either array.

UCS - san-port-channel over vpc

Similar Messages

Maybe you are looking for