First look at Ubiquiti mPower Pro power strip, home lab pricing, enterprise features, uncertain future

Posted by Paul Braren on Aug 17 2016 (updated on Sep 10 2016) in
  • Network
  • Power
  • HomeLab
  • HomeServer
  • Review
  • Ever wonder if there's an affordable power strip that you could add to your home network that would give you insight into the power usage of up to 8 devices, and allow remote power on/off operations?

    If you're willing to tolerate a bit of initial setup pain that requires a back-level Java and a firmware upgrade, you'll be rewarded with a solid built-in web server that will do nicely for basic watt use and power toggling, as seen pictured at top-right in the screenshot below. This will likely work well for my SuperServer/NUC demo coming up very soon at VMworld US.

    See also Ubiquiti's datasheet.

    Available at Amazon and Newegg.

    Just unboxed mine yesterday, used a Windows 10 VM with 2GB of RAM and 1 vCPU and Java 7 x86 just for initial setup, configuration, and firmware updating. Even without the firmware update, I could have just pointed any browser to the local IP address of the device and seen watt use and toggle power status at will, a LOT more affordably than with datacenter PDUs. See also possible reasons why it's priced pretty low.

    Video

    Ubiquiti Networks mPower Pro External Power Control Unit - unboxing.
    TinkerTry-look-at-mFi-mPower-PRO-ubnt
    Each system is at idle, running ESXi 6.0U2 off USB set to default Balanced profile, with one 2.5" Micron SSD for storage in each Supermicro SuperServer, and a Samsung 950 PRO M.2 NVMe in the Intel NUC6i7KYK.

    AUG 19 2016 Update

    I want to be sure you're fully aware of some valid reasons that are likely contributors to the low price of this product, given the software support is not long for this world. I had mentioned this in the original article with the quote below, but I wanted to also add this Reddit conversation too.

    The title of this article has now been modified appropriately, adding "uncertain future."

    2016-08-19_11-56-14

    I had upgraded the firmware to MF.v2.1.11 Build Number 1309 and plugged in ethernet, and the WiFi went away. Good. I'd much rather only use the wired connection anyway, especially at trade show demonstrations. But from some research, it would appear this auto-disable-WiFi feature was there in earlier firmware levels. So if you don't need fancy scheduling or cumulative stats, you can forget the controller, and just use the local IP address of the device.

    I do like to configure the configurable automatic browser display refresh rate, for the richest feature set view available through the mFi controller, where I can go all the way down to every 5 seconds. The unconfigurable-refresh direct web UI actually updates every second. Very handy for demonstrations and videos.


    Aug 26 2016 Update

    It appears you may be able to avoid Java/mFi Controller entirely, by doing some moderately fancy SSH work:


    Sep 10 2016 Update

    2016-09-10_10-52-30

    See also

    Refreshingly honest!

    • is MFI dead?

      Ubiquiti Employee UBNT-Robert
      UBNT-Robert
      Re: is MFI dead?
      Options
      ‎04-23-2016 08:41 AM

      Hi All -- I know an update is due from my side. Here is what I can say:

      1. Ubiquiti does have a long-term IOT commitment on-going
      2. mFi is not part of that long-term strategy
      3. We will continue to make mFi hardware, but SW will be placed on hold indefinitely

      I know a lot of you will be justifiably upset so I will provide some additional transparency:

      mFi development started in 2009 ahead of the "IOT" movement. The original goal was to provide a "canvas" to the market which allowed them to connect together through Internet Protocol any sensor, power outlet, switch, or serial port and create relationships between them from a centralized controller. This was going to be the encore to UniFi's breakout success.

      Unfortunately, the SW team I recruited to deliver the vision made some horrendous design choices and generally didn't know what they were doing. Even though early deliverables were really disappointing, I was afraid to lose time to market and kept on "doubling down" with the same team. In retrospect, I should have made the decision to replace the team and start over before launch, but I didn’t. And once we launched and marketed the platform, we were stuck in persistent maintenance/support mode without really solving the fundamental issues.

      From the java back-end selection to poor coding, poor installation UX, no remote connectivity, and hardware shortcomings, it has not been something we can easily fix. Our conclusion was we had to start fresh to deliver a product that can be a winner in the market. I know that might not be a consolation for many of you, but we will continue to produce and support mFi hardware.

      Robert