Microsoft reviews M365 resilience after Indian outage

Microsoft has ordered a review of its resilience regime for Microsoft 365 after finding an outage to the service in India was caused by "a physical fiber networking event" at a partner's edge datacenter location. The outage took place on July 15. In a preliminary incident report [PDF] Microsoft explained that the problem arose …

[An_Old_Dog]

No Telling

“a physical fiber networking event on the devices which manage the connection pathways from their Chennai Edge site to Microsoft’s Data Centers”. This could have been someone tripping over a cable in a rat's-nest-spilling-out-of-the-back-of-the-racks-and-onto-the-floor, and consequently yanking said cable out of a port, or someone accidentally moving a cable from the wrong port.

I regard the first possibility as negligence by the 3rd-party partner's managers having not spent the money and time for their engineers to set the cabling up properly (yes, I know fibre cables have bend-radius limits). The second ~might~ have been avoidable (were the cables and ports properly labelled, and those labels easily-visible?).

[Giles C]

Re: No Telling

It could also be the infamous road worker going through the cable in the road outside the data centre. A local cut fibre is always a pain to deal with, although you would have hoped on a connection this important (for both companies) that multiple redundant paths would be provisioned