Thread by @rnelson0

.@editingemily kicking off the keynote at #DevOpsDays Indy!

Her talk is titled Everything Is On Fire, and we just had a fire alarm. Seriously. It doesn’t get better than this! #DevOpsDays

Emily says we can learn a lot from other, older industries - including, of course, fire fighters. #DevOpsDays

The talk is going to explore how fire fighters work - before, during, and after a fire. #DevOpsDays

I should note, it was a false alarm, we are okay. It is, actually, fine :) #DevOpsDays

Firefighting is very regimented, like a military. But the industry changed rapidly about 50 years ago. Why? Funerals. Too damn many of them. #DevOpsDays

After serious forest fires in CA that killed many, a task force FIRESCOPE was created to investigate solutions. Interesting history to look up deeper lately. #DevOpsDays

FIRESCOPE persisted and after 9/11 turned into NIMS. Emily’s favorite part is that NIMS standardized the vocabulary of the industry. Something we don’t have in IT. #DevOpsDays

(This is something I personally harp on constantly. Yesterday we talked about shared context, and lack of a shared vocabulary can cause two people to think they have a shared context but talk completely past each other. We need to fix this.) #DevOpsDays

FIRESCOPE had 4 findings that I didn’t catch in time to document but are worth looking up yourself. Standardization was a huge part of it. #DevOpsDays

Rotate incident commanders. Include juniors - what better way to provide familiarity and confidence? You should be supporting every incident commander, anyway. #DevOpsDays

Incident Commanders (ICs) get to determine who should be involved! Both actively and who is on standby. Don’t need everyone involved but preparing standby people before they’re needed is helpful. #DevOpsDays

Emily suggests using a video call. In part because you can see glazed eyes which may be a sign of lack of context, or too long on shift in extended issues, not easily detected on audio/text only meetings. #DevOpsDays

IC needs to keep people focused. Esp ensure people aren’t worried about being fired.

Also need to take notes to record the event. Don’t need to be verbatim but inform post-incident reviews. Don’t weaponize them! #DevOpsDays

IC gets to communicate with leadership and maybe PR if external. Provide cover for teams, and ensure those members aren’t distracted by it.
Make sure executives aren’t on the call or have direct access to the incident responders, only you. #DevOpsDays

At the end of the incident, IC should send a final note to mgmt/PR, then have all responders take 5 minutes to jot down their independent notes. Schedule a review at that time, so it’s not forgotten. #DevOpsDays

Q/A/Comments now
C: I recommend if an executive hops on the call and overriding the IC, the IC should ask “are you taking over the response?” #DevOpsDays

Q: how do you get people to join a call, whether video or not? Brents like to just dive in
A: crowd sourced, lots of answers. Make sure IC doesn’t let it slide; call out “hey why is this person not on the call logged in?”, etc. #DevOpsDays

A (continued): “CTO said this is our process, you must comply” can go a long way. A benefit to formalizing incident review. #DevOpsDays

I really liked this talk by @editingemily, thank you! #DevOpsDays