Infrastructure:APC RackPDU

From Research
Revision as of 06:16, 18 August 2011 by Rmackenz (talk | contribs) (Update with new info, re-organize)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

We have a total of 15 APC Power Distribution Units. These allow for remote power monitoring and cycling for all connected devices.

Specifications

Device List

  • The following units are in the Research Server room, and are AP7900 units.
    • APC00 - Hood, Hercules, Navajo, Lancer, Tirpitz
    • APC01 - Tape Library, Musashi Raid1 Musashi Raid2
    • APC02 - Spitfire #1, Spitfire #2, Spitfire #3, Hurricane #1, Hurricane #2, Hurricane #3, Tapes and Screen, Avocent KVM
    • APC03 - Teleport, Bismarck, Musashi, Nelson, 48 port Dell switch
    • APC04 - HVI Lab, Echo, Slinky, Thunderbolt, Switch
    • APC05 - Clipper, Intendant

This list is kept in /home/projects/infrastructure/scripts/APC_Ports in a script called APCupdate.sh

Update the list at the top of this script, then run it to generate the config files. Then run the FTP script and the config files will be uploaded to each APC device, changing its config. The device will then reboot and come back in about a minute (No power interruption)


Details

Cluster

  • The following units are connected to Nebula, and are all MasterSwitch VM units.
  • Nebula units reachable from cluster head only.
    • IP range 192.168.254.NNN where NNN is the unit number.
    • APC01 - node001 to node022, 2 unlabeled outlets
    • APC02 - node023 to node044, 2 unlabeled outlets
    • APC03 - nodes057,060,063, 21 unlabelled outlets
    • APC04 - node064 to node084, 4 unlabeled outlets
    • APC05 - node085 to node106, 2 unlabeled outlets
    • APC06 - node109,112,115,118,121,124,127 and 17 unlabeled outlets

Monitoring

All data is fed into Nagios for monitoring. Units are tenth of an Amp (so 81 = 8.1A). Warning level is 9.0A, Critical level is 10.0A.