Microsoft extends 'outage mode' for Azure Active Directory to bake more resilience into cloudy services

Microsoft hopes to improve the resilience of its cloud services by extending an "outage mode" for Azure Active Directory to cover web as well as desktop applications. Azure Active Directory (AAD) is Microsoft's cloud directory which handles authentication for Office 365 and can be linked to on-premises Active Directory. …

[Anonymous Coward]

Msft on the back foot

The thing is, Msft made a huge mistake in letting Amazon run their cloud for years without any competition. Then, in a rush to catch up, they rolled out a global architecture that has all sorts of reliability issues. They are now trying to retrofit availability into their Architecture, but many of their higher-level services simply haven't caught up. The ones that have caught up now charge a premium for things like availability zones, that come as standard on other Clouds. As these services are being retrofitted to increase reliability, and are not designed from the outset, they are having to come up with 2nd rate solutions like the one mentioned in the article. They have a very poor track record of managing availability/capacity and frankly, I would be very concerned running any significant (resulting in loss of life) application on Azure.

[Pascal Monett]

Re: any significant (resulting in loss of life) application on Azure

I would be very concerned if any hospital ran any of its IT on any cloud whatsoever.

If your software can be responsible for loss of life, it is your responsability to ensure 100% uptime, with redundancies and failovers.

There is not Cloud that ensures that today, or even pretends to.

[Peter-Waterman1]

Re: any significant (resulting in loss of life) application on Azure

TBH - I don't see why an On-prem Data Center would be any more reliable than AWS which has a min of 3 isolated active-active Availability zones per region and then the ability for customers to run multi-region on top of that.

[Richard 12]

Re: any significant (resulting in loss of life) application on Azure

On-prem still works when a JCB digs in the wrong place.

[W.S.Gosset]

Re: any significant (resulting in loss of life) application on Azure

"radical soil surgery", in medical parlance

[Numen]

3 days?

"There is also a limitation in that authentications are only processed by the backup service if the user has already accessed an "app or resource" within the last three days"

So no more long weekends or longer holidays, like Thanksgiving?

[Richard 12]

Re: 3 days?

Presumably nobody at Microsoft needs to do work on Mondays.

[Howard Sway]

the root cause attributed to "outbound port exhaustion"

My advice is to just relax, it happens to everyone as they get older.

[deadlockvictim]

Outrage Mode

I just have to learn to handle Outrage Mode whenever Microsoft bombards me with e-mails telling me how wonderful Azure is and why we should be paying rent to Microsoft in perpetuity for every little thing we do.

Oh, and I'd like the certification exams for SQL Server back please.

[LateAgain]

Outage mode?

Brilliant idea !

It's off / failed / f*d but you can't complain?

[Anonymous Coward]

CosmosDB

It's not a surprise it has large issues. It was a MongoDB clone that got re-marketed into a One Database To Rule Them All when people were happy just using MongoDB for MongoDB things. Lots of changes of direction are bound to cause issues.

[Anonymous Coward]

Microsoft extends 'outage mode' for Azure...

Is that a nice way of saying that Azure stays up as long as a pensioner on valium does?

[Paul 87]

No idea why there isn't more effort put into a hybrid model or decentralized model, heck it could even be an actual, valid real world use for blockchain.