What Data Skew in Spark Really Looks Like — and Why It Quietly Breaks Your Jobs

There is a specific kind of failure that does not look like a failure at all. Continue reading on Medium »