Definitions

Question 1

RDWF

Answer 1

RDWF, or Rackdown Workflow, is the globally approved tool for rackdown recovery. It should be used during all rackdown events to recover customer service as quickly as possible.

Question 2

TOR

Answer 2

TOR is the acronym for Top of Rack switch. This is the network device that enables our hosts to be connected to the Amazon network.
Note: Even though these devices are called Top of Rack switches they can be located anywhere in a rack, so be careful and always assume there is more than one in each rack.

Question 3

Host

Answer 3

For a rackdown event, a host refers to a server that is connected to the affected TOR switch.

Question 4

EC2 Rack

Answer 4

EC2, or the Elastic Compute Cloud, allows users to rent virtual computers on which to run their own computer applications. EC2 racks support this service.

Question 5

Instance

Answer 5

An instance is a customer’s virtual machine. You can have multiple instances on a single server.

Question 6

First Responder

Answer 6

A First Responder is the first person (or persons) to respond when a rackdown event occurs. This includes Data Techs, EOTs, Chief Engineers, Queue Masters, etc.

Question 7

ATS

Answer 7

An ATS is an Automatic Transfer Switch. This is a device that automatically switches from one power input to another without dropping the power on the output, so that the equipment connected to the ATS will not detect any drop in power.

Question 8

rPDU

Answer 8

rPDU is the acronym for RackMount Power Distribution Unit. Hosts, TORs and other equipment on the racks get their power by connecting to these for units.

Question 9

RATS

Answer 9

RATS, or Rack Aligned Ticketing System, is our EC2 fleet health monitoring system. It is triggered whenever certain parameters are met that resemble a rackdown issue. This results in a Master TT being created for the Rackdown event. The Master TT is always placed in the local DCO queue.

Question 10

STORM Rack

Answer 10

The STORM Rack is a designated rack that contains the critical spare TORs that the first responder will need if required when replacing a TOR during a rackdown.

Question 11

Powershelf

Answer 11

Racks fitted with a Powershelf differ from legacy racks, in that they do not contain any rPDU’s. Instead, the powershelves give a direct DC supply to components in the rack. This removes the need for these components to contain PSU’s.

Note: It should be noted that different powershelves power different components. For example, a Waipoua powershelf might only supply the hosts in a rack, whereas a Who Made Who (WMW) powershelf will supply all components in the rack, the hosts, console and TOR switch.
The following training is available on Knet:
Data Center Safety: Who Made Who (WMW) Safety Awareness
Waipoua Power Shelf ATS Replacement