Operations in a Distributed World - The Practice of Cloud System Administration - page 166

7.4.1 Communication Mechanisms
Chat rooms are commonly used for staying in touch throughout the day. Chat room tran-
scripts should be stored and accessible so people can read what they may have missed.
There are many chat room “bots” (software robots that join the chat room and provide ser-
vices) that can provide transcription services, pass messages to offline members, announce
when oncall shifts change, and broadcast any alerts generated by the monitoring system.
Some bots provide entertainment: At Google, a bot keeps track of who has received the
most virtual high-fives. At Stack Exchange, a bot notices if anyone types the phrase “not
my fault” and responds by selecting a random person from the room and announcing this
person has been randomly designated to be blamed.
Higher-bandwidth communication systems include voice and video systems as well as
screen sharing applications. The higher the bandwidth, the better the fidelity ofcommunic-
ation that can be achieved. Text-chat is not good at conveying emotions, while voice and
video can. Always switch to higher-fidelity communication systems when conveying emo-
tions is more important, especially when an intense or heated debate starts.
The communication medium with the highest fidelity is the in-person meeting. Virtual
teams greatly benefit fromperiodic in-personmeetings. Everyonetravels tothesameplace
for a few days of meetings that focus on long-term planning, team building, and other is-
sues that cannot be solved online.
7.4.2 Communication Policies
Many teams establish a communication agreement that clarifies which methods will be
used in which situations. For example, a common agreement is that chat rooms will be the
primary communication channel but only for ephemeral discussions. If a decision is made
in the chat room or an announcement needs to be made, it will be broadcast via email.
Email is for information that needs to carry across oncall shifts or day boundaries. An-
nouncements with lasting effects, such as major policies or design decisions, need to be
recorded in the team wiki or other document system (and the creation of said document
needstobeannouncedviaemail).Establishingthischat-email-documentparadigmcango
a long way in reducing communication problems.
7.5 Summary
Operations is different from typical enterprise IT because it is focused on a particular ser-
vice or group of services and because it has more demanding uptime requirements.
There is a tension between the operations team's desire for stability and the developers'
desiretogetnewcodeintoproduction.Therearemanywaystoreachabalance.Mostways