Ask the Expert: Demystifying Unified Computing System (UCS) Interfaces for troubleshooting

[ciscomoderator]

Ask your questions following the Live webcast, March 17, 2015 ending March 25th

Ever wonder what VFC, VETH, VIF and HIF are in UCS and which path your packets are taking?   UCS infrastructure has several virtual components and this makes it challenging to troubleshoot but it is critical to understand. In this Webcast, Niles will discuss UCS interfaces and how packets travels from the UCS server. This includes what happens when a packet leaves a server from the vnic until it traverses the Fabric interconnect.  Once you understand the data path, you will be able to troubleshoot the Unified Computing System

Niles Pyelshak is a Customer Support Engineer for Cisco TAC works in Server Virtualization Team.  She has several years of networking and virtualization experience.  She recently passed her CCIE Certification in Data Center (#44608). As a Network Engineer, Niles was accustomed to getting on the command line interface (CLI) to configure and build her network. In server virtualization it is a little different. She is always amazed at how servers are mostly managed by Graphical User Interface(GUI) interfaces and CLI.  One can not underestimate the power of both interfaces and in this Webcast, Niles will leverage both GUI and CLI to demonstrate the UCS life of a packet.

Find other  https://supportforums.cisco.com/expert-corner/knowledge-sharing.

**Ratings Encourage Participation! **
Please be sure to rate the Answers to Questions

   

[tenaro.gusatu.novici]

Hi Niles,

thanks for providing us this opportunity to learn more about UCS.

I have one "hard" question to begin with: what will be your favorite CLI troubleshooting commands? I understand this is not very precise question but based on your experience as TAC engineer maybe you can remember 3-4 commands that will reveal things otherwise not easily seen in GUI.

In addition, if it is not out of the scope, is there 16Gb FC interface available on Fabric Interconnect?

Thanks,
Tenaro

[tenaro.gusatu.novici]

I posted question four days ago and still nothing. I tried to contact ciscomoderator and that redirected me to "contact us" form in the upper right corner but still no reply. Anybody, any ideas how to reach real people behind aliases?

[Bruce Heimlich]

Hi Tenaro,

Let me reach out to Niles and let her know about the pending questions and I am sure that she will be glad to answer them all!

Thanks Regards,

-Bruce

Update: I reached out to her but it looks like she is out of the office today. That probably explains the lack of response.

[Niles Pyelshak]

Thanks Bruce!  Cheers!

[Lisa Latour]

Hi Tenaro - Sorry about the "alias" - we often use that ciscomoderator alias so that the focus is on the topic or the expert and not the poster. We do monitor as we can, and we do respond on the contact form. Thank you for being part of the Cisco Support Community. Your participate helps others!

[Niles Pyelshak]

Hi Tenaro,

Thanks for posting a query!  

Did you get a chance to look at the presentation files?  That will give you some of my favorite UCS cli commands that I use when I'm determining which path my compute machines are using.  If you have any specific queries, I'd be happy to address.

One CLI command that I use religiously is "Show cluster state".   Why?  One of the greatest advantage of UCS is that if best practices are followed, the UCS domain will be highly available.  So I always check cluster state status to ensure that both "HA is ready" on both Fabric Interconnects.

My apologies for not getting back to you sooner.  Are you aware that, technically, we are not allowed to answer queries until after the Webcast event date?  If not, pardon the miscommunication.

In regards to 16Gb FC, yes...it's a bit off topic but that quite alright.  I'm here to help.   Unfortunately, I do not know the correct answer for this.  I typically see FC link aggregation with the use of FC port-channels in deployments where bandwidth is a concern.  You can bundle up to 16 native FC links in a port channel.  I'll have to check with the TME and get back to you.

Again thanks,

/Niles

[Niles Pyelshak]

Hi Tenaro,

I've confirmed that 16G FC interfaces is currently scheduled for 3rd Gen FI road mapped for Q4CY15.

There is your answer!  I'm happy to help!

/Niles

[Telco Geek]

Is it possible to delete Audit Log on UCS-M? If yes, can you explain how to do it?

Thank you,
Jose

[Niles Pyelshak]

Hi Jose,

Thanks for you question.

You can only view, export, print or refresh the audit logs in the latest UCSM release.  Deleting the audit log is not currently an option.

http://www.cisco.com/c/en/us/td/docs/unified_computing/ucs/sw/gui/config/guide/2-2/b_UCSM_GUI_System_Monitoring_Guide_2_2/b_UCSM_GUI_System_Monitoring_Guide_2_2_chapter_0110.html#task_5AE365A455F044D5903EF932DADD9CFD

HTH,

/Niles

[Doug Woznicki]

Thanks for the great info Niles.  I use MS Hyper-V 2012 R2.  I have a couple new 6248 FI's running 2.2.3c.  I would like to know if the is a best practice guide or any good information for configuring UCS all the way down to the servers for Hyper-V.  I have B200M2, M3 and M4, B230 M2, C240 M3S and C220M3S in the environment

There is an older document but it doesn't take into consideration the new technologies

Thanks

Doug

[Niles Pyelshak]

Doug,

Thanks for coming to the Webcast!  Glad it was helpful!

Here is the recommended practices for Hyper-V deployment:

http://www.cisco.com/c/dam/en/us/solutions/collateral/borderless-networks/advanced-services/Whitepaper_C11-730725.pdf

Here is a config sample for VM-Fex implementation:

http://www.cisco.com/c/en/us/support/docs/servers-unified-computing/ucs-manager/116374-configure-vmfex-00.html

You can also deploy Windows 2012 in NPIV mode which allows all the guest Virtual Machines that run on a single server to use a single adapter while the VMs still maintain its' own separated protected storage.

Check it out below if you are interested.

http://www.cisco.com/c/en/us/support/docs/servers-unified-computing/unified-computing-system/116471-configure-npiv-00.html

HTH,

/Niles

[Doug Woznicki]

Thanks Niles.  The links provided were great. I will be looking to implement what fits as soon a feasible.

[Francisco Macias]

hello - How does my vnic determine what uplinks it’s pinned out in EHM?

[Niles Pyelshak]

Hi Francisco,

Thanks for your query!

Q.  How does my vnic determine what uplinks it’s pinned out in EHM?

A.  You have 2 choices once your vNIC in service profile is specified to use a certain Fabric Interconnect(A or B)

1.  Dynamic server to Uplink pinning that is setup by the UCS system automatically(single uplink  port or port-channel).  VLAN membership is taken into consideration here.
2.  Static Pinning with the use of LAN pin groups.  This could doesn't scale well and could be a  management nightmare.

Checkout this whitepaper:

http://www.cisco.com/c/en/us/solutions/collateral/data-center-virtualization/unified-computing/whitepaper_c11-701962.html

COnfiguring LAN pinned groups - Chapter 18

http://www.cisco.com/c/en/us/td/docs/unified_computing/ucs/sw/gui/config/guide/2-2/b_UCSM_GUI_Configuration_Guide_2_2.pdf

You may also be interested in disjoint Layer 2:

http://www.cisco.com/c/en/us/solutions/collateral/data-center-virtualization/unified-computing/white_paper_c11-692008.html

HTH,

Niles