21Spark executors sit Pending for hours because the default scheduler fragments them across nodes. Can you bring in a different scheduler?▼mediumNewDatabricksNVIDIAFlipkart◆ premiumMultiple schedulers are supported, rarely understood, and dangerous in a specific way. The strong answer names the deadlock that gang scheduling fixes and the capacity race it can open.Open full answer →
06Your batch fleet runs on spot and keeps losing nodes mid-job. How do you make interruptions survivable?▼mediumNewAmazon & AWSFlipkartDatadogunlockedEveryone quotes the discount. Fewer candidates can describe what happens when a notice arrives late or a worker disappears without completing shutdown and why some workloads shrug it off while others lose hours of compute.Open full answer →