Your weekly Azure outage

The cloud is no different than on-prem in terms of having to plan for resilience during outages..
 
Kosmik said:
My databricks clusters are complaining :(
Click to expand...
Howdafuq do you expect bricks to stay floating up in a cloud? Especially when there's a bunch of them glued together?
 
Sinbad said:
We had that little storm yesterday.
Got a bit damp this side, our infrastructure handled it ok. Why are you using toilet paper for your fibre optic comms? :D
Click to expand...

That little storm passed right over our town today, code red, alarms sounding the works.

My home fibre connection had no problems, that MS data center though - uber kak
 
WAslayer said:
The cloud is no different than on-prem in terms of having to plan for resilience during outages..
Click to expand...

On prem "cloud" is busy loading in the org. Think they've had enough of this - so much cost, so many issues and then our legacy on prem stuff chugs along.
 
Hamster said:
On prem "cloud" is busy loading in the org. Think they've had enough of this - so much cost, so many issues and then our legacy on prem stuff chugs along.
Click to expand...

SaaS is better than both :p (Where possible). Make it someone elses problem.
 
WAslayer said:
Until you find out your SaaS provider is on Azure and affected by these outages and you are affected anyway...
Click to expand...

Ideally, they implement HA and failover to keep to their SLA targets and keep all their customers happy, plus they invest way more in the skills and strategy behind hosting vs a company whose core competency is something else. But fair point nonetheless, could and surely does happen :)
 
Bobbin said:
Ideally, they implement HA and failover to keep to their SLA targets and keep all their customers happy, plus they invest way more in the skills needed to host vs a company whose core competency is something else. But fair point nonetheless, could and surely does happen :)
Click to expand...
Going straight back to my first post here.. moving to the cloud does not mean you don't need HA and failover anymore.. so money companies still don't realise this or don't want to spend the money for it.. but they are quick to complain about an outage..
 
Bobbin said:
SaaS is better than both :p (Where possible). Make it someone elses problem.
Click to expand...

Dear valued client,

We sincerely apologise for the outage on our platform that serves a couple of million of clients.

A while back we made the decision to offload critical functionality (including authentication) to a third party provider for cost cutting measures - savings we passed onto you our deserving clients.

What we unfortunately didn't foresee was that this approach - usually reserved for dinky toy projects - would become a single point of failure. The fact that our clients lost money because of this hurts us, but not as much as the financial and reputation damage we suffered.

Please accept our heartfelt apologies.

Sincerely,
CEO posting on LinkedIn about his startup scaling up unicorns and doing things right!
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X