Power BI is a SaaS product. Doesn't that mean that Microsoft handles all the operational concerns for us? Do I even need to worry about "Power BI operations" in my organization? To answer these questions, we need to understand what operational concerns are relevant to Power BI. In answering this question, we also define our own scope and mission here at Power BI Ops.
The metaphor of space travel is a good one. Launches are exciting. Huge crowds will travel for miles and wait patiently even for days to witness a launch in Cape Canaveral. They encompass a full plot complete with heroes, suspense, a climax and successful resolution. A launch is the culmination of months and years of planning and engineering.
We tend to prize launches similarly, though perhaps with less public acclaim, in software and analytics endeavors.
As exciting as it may be, the launch is not the purpose of a mission. In fact, it is only one necessary step to completing the mission objectives. Although the launch tends to draw more attention than the mission that follows, the mission is where the objective is fulfilled.
Similarly in software, the purpose of all the analysis, engineering work, and launch is to serve some business use case.
The mission is why we launch. The use case is the value of the software. "Operations" is the name for activities that support the mission, the use case. For NASA, operations is embodied in Houston of, "We have a problem," fame. For most organizations, it is their IT department that is responsible for operating tech solutions.
With that definition, Power BI operations can be defined as everything that needs to happen to continue to get business value from Power BI after we do the hard work of developing and launching a solution.
Traditionally, software operations has included deploying and monitoring things such as:
- Compute resources
- Storage
- Networking
- Logs
- Configuration management
- Resource utilization
- Backups
- Availability
- Disaster recovery
- Patching and deployment
Operators must answer questions such as, "What is happening?" and "What has changed in the environment?". Operations concerns itself with the entire life cycle of Power BI content.
Power BI is a SaaS, though, so what does that provide for us operationally? We get high-availability built in -- no more racking and stacking machines in data centers for you, Power BI Admin! We get a fully managed and virtualized infrastructure that is a very powerful black box to us. We also get some notifications and administrative controls built in.
But that still leaves a lot up to you when deploying, administering, governing, and operating Power BI at scale.
- Logs: Power BI captures them, but doesn't do anything for us with them.
- Configuration management: we own all the settings in our tenant.
- Resource utilization: this especially impacts those who manage dedicated capacities.
- Backups & disaster recovery: content that is pushed into the Power BI Service can be lost and changes cannot be rolled back.
- Deployments: we still manage the deployments and lifecycle of all content.
These are important questions. Concretely, these can take forms such as below, and so many more:
- What are all of the refresh failures that have occurred in the last week?
- What dataset refreshes take the longest?
- What content is unused in the organization?
- Who is publishing and editing content?
- Who can see our sensitive data sources?
- Can we restore the version of some content that someone accidentally deleted?
- What workload is consuming all of the resources for my Premium capacity?
With this site, we will be walking you through the resources that you have available to answer these important questions. We want to make sure you have access to all the tools available to you as a Power BI administrator and operator. Sign up for our newsletter and follow us to keep up with everything you need to operate Power BI effectively.
