Calculate downtime with days of different length

[Anonymous]

I need to calculate the downtime of various critical systems (think telephony, website, etc.):

• Per system
• Per business process affected
• Overall for the organization
• Per month and YTD.

Why it's complicated: 

• Some systems need to be up 24/7 (website, for example) and others during specific periods (operating hours): the sales DB might need to be up 8:00 to 20:00 Mon-Fri, 9:00 to 15:00 Saturday and closed on Sundays whereas telephony and intranet would be 7:00 - 22:00 7 days/week, etc.
• Some downtime events might be over more than one calendar day and those days may have different operating hours.
• Some events affect more than one system, and those systems might have different operating hours.
• Some events affect more than one business process.
• Schedules might change during the year

Here's a sample of data I have at this point: (I do have a date dimension table and can create a time dimension table as well)

Outages

Systems to Business process mapping

Systems operating hours

Everything I found so far is for simple, fixed hours workdays with the only twist being weekends/holidays.

Anyone has an idea how I can achieve this?

[v-jayw-msft]

Hi @Anonymous ,

If I understand you correctly, I think the entry point is to divide the rows which downtime more than one day into day-by-day rows as below:

BTW, could you please share some sample to me if you don't have any Confidential Information?

Best Regards,

Jay

[Anonymous]

Hi @v-jayw-msft 

I don't have access to online file storage from work. 

Otherwise, the screenshots are from a dummy file that doesn't have confidential information: I invented the names, dates, schedules, etc. but the structure is identical. I did save it just in case.