Virtual Disconnect: Migrating from HP BladeSystem to Cisco UCS

Aug. 26th, 2010

Now in our second year of shipping Cisco Unified Computing System (UCS) blade servers, it seems that most of our 1700+ UCS customers are migrating to Cisco blades from HP blades. Two teammates of mine, Jeff Allen (Twitter) and Doron Chosnek (Twitter), have developed a process to help these customers migrate from HP blade servers to Cisco UCS blade servers in a matter of minutes. Their migration process, based on publicly available documentation from HP and Cisco, allows a UCS customer to move the “server identity” of the retired HP blade server to the newly deployed Cisco UCS blade server. For a discussion of what the “server identity” is, see this previous blog entitled “The State of Statelessness: Cisco UCS vs. HP Virtual Connect“.

Doron and Jeff accomplish this HP-to-Cisco migration by extracting the server identity information out of HP Virtual Connect and then importing that server identity information into Cisco UCS Manager. Fundamentally, they export an HP Virtual Connect Server Profile and import it into Cisco UCS Manager as a Cisco UCS Service Profile.

Server identity migration can be done seamlessly from HP to Cisco because Cisco’s UCS Service Profile is a superset of HP’s Virtual Connect Server Profile. Cisco manages ~96 server identity settings while HP only manages ~12 server identity settings. The reverse process, migrating Cisco-to-HP, would be difficult (if not impossible), since HP supports so few server identity settings in comparison to Cisco. I’ve provided an updated Cisco UCS vs. HP BladeSystem server identity comparison below. See Table 1.

In the following video, Doron and Jeff demonstrate their process by migrating a workload from an HP BL460c to a Cisco B200 M1 server. In their setup, both the HP blade and the Cisco blade have access to the same Ethernet and Fibre Channel networks. The soon-to-be-retired HP blade is booting from SAN. During the migration, WWN and MAC addresses, boot from SAN target and LUN number, etc. are taken from the HP server and given to the Cisco server. Since the FC network and storage controller see the same server identity on the new UCS blade, this allows the Cisco blade to immediately boot from the same FC LUN that the retired HP server was booting from. The result for the customer is an upgraded server running the same instance of the operating system with no changes required on the LAN or SAN.

UCS Documentation References:
Overview of Cisco UCS Manager
Cisco UCS Manager CLI Command Reference Guide

____

Table 1: Comparison Chart Showing User Interfaces Required To Configure A Server (click table to enlarge)

Aug 26th, 2010 | Posted in Cisco UCS
Tags:
  • Ddelfino

    Doron and Jeff Great Job!!!! … ddelfino

  • Jeremy Waldrop

    Very cool guys, nice job. Any chance I can get the perl script?

  • Jeremy Waldrop

    How did you account for the different HBA driver? Did the HP blade use a Qlogic or Emulex HBA?

  • Jeff Allen

    Conveniently, both servers use the same exact Qlogic driver. Because this was Windows, we had to add the Cisco PCI DeviceID to the registry prior to the move. This is a really simple edit and is not needed on other operating systems. Let me know if you want more info on this – I can expound a lot more, but it gets boring for most people.

  • http://twitter.com/dchosnek Doron Chosnek

    Jeremy, the script will be posted on the Cisco Developer Network (CDN) where there are already quite a few sample scripts for interfacing with Cisco's UCS. You can access CDN with this link: http://developer.cisco.com. If you have any specific questions or need some ideas for your environment, let us know.

    (remember that scripts posted to CDN like the script demonstrated here are not just examples or starting points for you)

  • Marotta

    Awesome ! Really great stuff – I especially find this useful when HP customers ask: “I currently have HP blades, but I want UCS, how do I migrate easily ?”

    Thank you, thank you, thank you for this wonderful video (and Table 1) !

    – Mike

  • http://twitter.com/nuocbap Quoc Binh Ngo

    Thank you very much,
    I understand why UCS profile is better.

  • http://twitter.com/joechin99 joechin99

    We are conducting a UCS evaluation. We have observed that when running Prime95 (64-bit default test) on a B200 and a B250 blades, the processors run at much higher reported temperature than equivalent HP blades (BL460). On the UCS, the CPUs were idle around 40 C, after 10
    minutes of testing, the reported temperature peaked at around 80 C, before levelling off around 75 C for an hour long run of the test. We repeated the same tests on two HP blades with equivalent processors. The starting idle temperature was around 40 C, and peaked at around 44 C.

    The UCS blades have the latest baffles installed. The Cisco folks were able to reproduce the same observation in their lab systems. They told us this is within operating spec for the processors, so nothing to worry about. But many people in our organization are still concerned at this difference.

    For those who have (access to) both Cisco and HP blade systems, do you observe the same temperature difference? How can this be explained? Do Cisco and HP use different instrumentation and/or methodology for acquiring and reporting processor temperature?

    • http://twitter.com/joechin99 joechin99

      An update …

      One of my team members came across a program called RealTemp which reads the temperature from the CPU’s directly. The software uses Prime95 to heat the CPU’s up and reports on each core temp along with showing their “tjmax” values, basically the values on the chip before thermal throttling kicks in. It appears to be ~100′C on these chips. He ran the exact same tests on both the UCS B200 blade and a BL460c G6 and it turns out that the UCS is reporting the actual CPU temps, while HP is NOT. The HP results from RealTemp were ~5-7′C cooler than the UCS, NOT ~30-40′C cooler! Bottom line, both blades run at about the same temps.

      So for the record, the UCS blade DOES NOT run hotter than the HP blade :)

      Also, here’s a report that provides detailed comparison of the power consumption characteristics between the B200 and BL460c G6 http://www.principledtechnologies.com/clients/reports/Cisco/UCSPower0310.pdf

  • Kris_tm

    Well joechin, How much did Cisco pay you for publishing the one sided report ?

    • http://www.mseanmcgee.com M. Sean McGee

      Hi Kris_tm,
      Three comments:
      1. I’ll risk sounding like a broken record – let’s please disclose who we work for when commenting. I’ll assume you’re an HP employee since you’re posting from 15.203.233.79 (http://cqcounter.com/whois/?query=15.203.233.79).

      2. Classy misdirected comment. Joechin has nothing to do with Principled Technologies, the company Cisco commissioned to do the testing. Joechin simply referenced the testing report. Joechin is an independent consultant.

      3. Do you know who the Tolly Group is? If not, you should. That’s who your company pays “to conduct custom testing designed to validate the vendor’s marketing message.”
      http://www.tolly.com/tollycertified.aspx

      Pot trying to call the kettle black?

      Thanks for stopping by,
      =sean

  • http://viettelonline.com/ ThuanTM

    so great! thank your infomation so much