Powerbi can easily handly 1 mill rows as it's a highly compressed column store data structure.
Load times will vary depending on the complexity of your data and the amount of differnt values. I recommend spitting date and times to reduce the different values and help the compression.
You can have 2 queries with current and history data and then merge the results. You can disable refresh on the history and update only when you change the boundary between current and history..
Direct query is also an option that only queries your data when you need it. I don't think MySql support direct query at the moment.
If you are using the power bi service and on premium then you can use incremental loading feature.
https://powerbi.microsoft.com/en-us/blog/incremental-refresh-in-power-bi-premium/
There is also aggregation data so you can have summaries rather than details for all your data and direct query can then get small data sets for details if needed.
https://powerbi.microsoft.com/en-us/blog/aggregations-for-petabyte-scale-bi-available-in-the-power-b...
Other Options are Azure Analysis services or on prem Cubes that can be loaded incrementally.
