[ano-ther]

This happens often enough (even MS itself broke their internal Exchange servers).

And every story seems to end with admins having to improvise. Am curious: (why) isn’t there a “kill reply-all chain” button as a feature?

(The article explains that this didn’t work for RIM because of BB’s architecture, but for Exchange?)

[eddieroger]

In theory, mail should be stateless, so what would they use to do that? Some clients and servers understand headers for thread identifiers, and then it would be possible to (in theory) zap that id, but all it takes is one client not supporting it for that to be out the window. The article kind of points this out - Exchange had a mitigation, but the clients got the notification and were able to reply because by then the mail was on the client (pushed).

[wiredfool]

The search term for MS for is bedlam 3.

[RandallBrown]

I worked at MS in 2012 and we had another email storm that shut down almost everything. I remember not being able to do any work at all because the network was so slow I couldn't even edit files (the internal source control system required getting a lock on the file to be able to edit them).

[philipwhiuk]

The standard fix is to massively limit who can send emails to DLs - which is an Exchange config option.