Smells like enterprise

by Volker Weber


I asked: "Do Amazon, Google and Microsoft also have these scheduled cloud outages?" Clemens Vasters from Microsoft simply answered: "No". And then elaborated:


My opinion: Better Enterprise than paying the price for patience during the day.

Sascha Troll, 2016-01-21

Your logic is wrong, Sascha, Clemens has it completely right: Needing scheduled downtime is a sign of higher probability of outages, not lower.

Stefan Tilkov, 2016-01-22

Remember the good old days when eBay was down every Friday noon for scheduled maintenance?

Daniel Haferkorn, 2016-01-22

No, I don't. Incredible!

Volker Weber, 2016-01-22

MS doesnt have that, but they have some sort of "service degradiation" at their enterprise cloud exchange thingie almost every other day.

Sascha Reissner, 2016-01-22

Martin Funk, 2016-01-22

A big IT shop (for a fortune 500) where I worked some 15 years ago had the policy to shut down every server saturday night at midnight and restart it at 1 am . This was mainly to ensure that all applications were build to support a shutdown and that there were no hidden state dependencies in the multiple connections they had. (probably > 100 servers).

I liked the policy but there was no system facing consumers at the time and this may not be a possibly policy when moving lots of stuff to cloud based services and when dealing with consumers.

Our current ERP development system (the software itself and the linux server) has an uptime of > 3 years.


Michael Diehl, 2016-01-22

In my view, the issue with Office365 is less a occasional hiccup, but the lack of any credible support. I cannot speak for companies with large accounts and thousands of users, but for SMEs you are left with a first level support that stops talking to you and closes tickets as "resolved" as soon as they run out of ideas, which happens quickly for all but the simple issues documented (and found) in their knowledge base. Paid support is only available with enterprise agreements, and local or outside contracted IT expertise is worthless as large parts of the service are a black box that cannot be properly diagnosed.

People might think of data protection and NSA when they reject cloud services, but for all but large accounts support in case of need seems to be the by far bigger issue. The IT dept might annoy you with scheduled service outages, but it is the same people you can call and that have access to their systems if things don't work. Just having somebody "apologizing for any inconvenience" does not even make you feel better.

Peter Daum, 2016-01-22

Sascha's right.

Download the O365 admin app onto your smartphone and watch those "Service Degraded" notifications come through.

Stephen Bailey, 2016-01-22


Ingo Seifert, 2016-01-22

The pros do something like this:

Here's IBM's:

Bruce Elgort, 2016-01-22

If I understand the comments correctly, IBM has planned cloud service outages but never any service degradations.

Volker Weber, 2016-01-22

@Volker, you need cloud customers for that.

Bruce Elgort, 2016-01-22

Cloud based service maintenance seems not so unusual:

Torben Volkmann, 2016-01-22

Are they hosting on SoftLayer?

Volker Weber, 2016-01-22

If they haven’t changed anything, they don’t:

Torben Volkmann, 2016-01-23

Old archive pages

I explain difficult concepts in simple ways. For free, and for money. Clue procurement and bullshit detection.


Paypal vowe