Skip to content

IT-4799/IT Power benchmark

Requirements

The power consumption of the proposed equipment shall be measured according to the instructions of this document. The results of the measurements and the methodology used to obtain them shall be reported to CERN in the Technical Questionnaire (Annex A).

The power consumption of the proposed equipment shall be measured with one of the following ways:

  1. Using a high-precision power analyzer in the primary AC circuit. In particular, the power analyzer should have a maximum tolerance of 3 permille.
  2. Using another hardware and/or software method which can be equivelant to option 1.

All tests shall be performed with the required Red Hat Enterprise Linux 9.5 installation.

Storage expansion unit power consumption

  1. As described in requirements, prepare the chosen method to measure the power consumption
  2. Power on the storage expansion unit and wait for it to initialize completely. The load average as indicated by uptime should be near 0.
  3. Measure the Real (W) and Apparent Power (VA) and Power Factor for 20 minutes and calculate the average value. This is the 'idle' power consumption of the storage expansion unit.
  4. Start FIO: fio --direct=1 --ioengine=libaio --group_reporting --rw=readwrite --output=storage-expansion-readwrite.log --bs=1m --time_based --runtime=30m --name=$DISKS where $DISKS shall be the list of block devices belonging to the storage expansion unit.
    As an example: /dev/sd{a..x} (to be adjusted to the actual configuration of the system).
    No other block device shall be included in the list.

  5. Wait for 10 minutes for the power consumption to stabilize.

  6. Start measuring the Real (W) and Apparent Power (VA) and Power Factor for 20 minutes.
  7. The average power consumption over that 20 minute period is the 'loaded' power consumption of the storage expansion unit.

Both the idle and loaded power consumption values shall be reported in the Technical Questionnaire.