If you're making a change where you're unsure if someone can perform a physical recovery, you're flying without a chute. It works most of the time until it doesn't.
(infra/physical colocation/hosting/sysadmin in a previous life for ~20 years)
EDIT: @mprovost (HN is throttling me, so I have to reply to your comment here):
Absolutely. It's a combination of experience, context, and wisdom. I myself have made changes that I believed were trivial but, because I lacked context, data loss or an outage occurred. These are painful lessons to learn. A culture of psychological safety and commitment to continual process improvement are key, with shared responsibility between contributors and the org. You are going to fail, but you shouldn't get cold sweats when you do as long as you did your best to derisk.
If you do the cost benefit of allowing people to work from home vs the cost of this one outage, I bet the outage is more expensive.
I mean if it happens naturally (say, meeting up after work for a drink) then sure but I don’t do it as a matter of course.
That said, I don't know what's so wrong with telephones.
Do you have all your coworkers' personal phone numbers?
My manager has my phone number, as does HR, and I imagine this is a standard situation at nearly every company.
The point was to make sure that, even if all of our infra including third-party IP comms and chat providers was hard down, we'd still be able to talk to one another and figure out what to do about it.
For an organization on Facebook's scale to overlook something so fundamental implies lots of things about the collective blind spots of its engineering and operational culture. None of those things make me less glad I so assiduously avoid even using any of its products, to say nothing of relying on them - at least to the extent that I'm able to avoid it, given their apparent ability to break half the anglophone Internet more or less at whim.
Mostly I was hoping to hear some amateur enthusiasm gloating about invisible light resilience.
Where I live there is a General Motors amateur radio club. The club is not owned or run by GM, but GM sponsors it by allowing repeater equipment on top of their building, and presumably monetary and other donations over the decade. Many of the members are current and former GM employees but you don't have to be affiliated with GM in any way (or even have an amateur radio license) in order to join.
There are lots of amateur radio clubs associated with a company for sponsorship/community outreach reasons, I assume the Facebook one is the same. (And I have to assume because they don't seem to have a website that I could find.)
To reply to another comment in this thread, you can break nearly all FCC rules if you need help in a life-threatening situation. Facebook being down was NOT a life-threatening event as defined by the FCC, even if their services being unavailable indirectly lead to loss of life.
https://www.ecfr.gov/cgi-bin/text-idx?SID=1a361a6eb3d1594e6a...
Part of a SOC2 audit is demonstrating that you have a good BCP, and part of a good BCP is demonstrating that you have fallback communication mechanisms available. So, either their BCP was out of date, or possibly they just ignored it during the moment of crisis.
And you know what's excellent for that ... literal word of mouth. I'm back in the office with 5 of my coworkers right now, and when we have issues we can just talk about them, zero technology needed.
It may also mean that you need more decentralization. They put all their eggs in one basket, so they took down Instagram, Messenger, and WhatsApp as well as Facebook, as well as their internal network, their building security, everything. Seems that they need to look at redesign to create more resiliency even if some efficiency is lost, so they don't lose everything again at some point.
I would expect the value of dogfooding to be massively larger than the cost of internal losses in cases like this. But it's good to remember to have a plan B.
To understand that the fix is a change to equipment in a data center, and on top of that what that change needs to be, how to make that change safely (though in this situation it's hard to imagine making things worse).
Everything else I’m reading says that January 2022 was always the planned return-to-office date ( https://www.cnbc.com/2021/08/12/facebook-delays-return-to-of... )
The claim that this was related to the recent outage doesn’t match all of the pre-outage articles that also point to the January 2022 date.
Unless someone has evidence that this is a recent change, I think this is a case of heavy editorialization by a journalist eager to capitalize on the combined anger toward Facebook and return-to-office plans.