Add memory consumption tracking to df execution report

[norberttech]

Right now when analyze option is configured and passed to df()->run() dataframe following statistics are goign to be collected:

• execution time (hrtime)
• total processed rows (int)
• schema (optionally collected, when configured)
• column statistics (optionally collected when configured)

Another interesting metric that I believe we can collect next to hrtime is memory consumption through memory_get_usage(true)

That will simplify monitoring of pipelines