Don’t get too comfortable with SLA’s

I have been working alot in Azure lately…more so than Office 365, but Office 365 still holds my heart. The other day there was a 5 hour outage with Visual Studio Team Services and Azure Active Directory. I monitored the situation, communicated to my colleagues and patiently waited it out. That night I got the health report that everything was fixed, so the next day I filed an SLA violation request.

It’s important to know where and how to determine SLA’s for Microsoft services, and they keep that on the volume licensing site.

Visual studio itself has a 99.9% uptime SLA. Remember this is just an SLA and not a guarantee, if this is violated you get a 10% refund, and if it falls below 99% it’s a 25% refund

The calculation they use is simple enough

Monthly uptime % = (Max Avail Minutes – Downtime / Max Avail Minutes)

In my case, 5 hours x 60 minutes = 300

31x24x60 – 300 / 31x24x60

44340 / 44640 = %99.32

That puts me at 10%. At first I was excited. I get 10% of around $1000 back…yay me $100.

Wait a second…I heard from around 5 development teams that were down with a total of 20 developers if I am being conservative. Assuming not all of them were working in that moment in VSTS.

I was also conservative with their hourly rates since I don’t know what they make. Say they make $20/hour which is extremely conservative since developers more than likely make more than 40k/year.

20 developers x 20/hour x 5 hours = $2000.

Now SLA’s don’t seem like they are such a good deal. Like I said before, don’t fall in love with SLA’s I am still conservatively $1900 short and if I am being generous it looks more like $7400. My $100 refund doesn’t look so good now.


  1. I’m not that much of a online reader to be honest but your sites
    really nice, keep it up! I’ll go ahead and bookmark your site to come back in the future.
    Many thanks

  2. It is an excellent site!A lot of helpful information and handy suggestions, bless you =)

Leave a comment

Your email address will not be published.