[FLINK-32254] FineGrainedSlotManager may not allocate enough taskmangers if maxSlotNum is configured

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.16.0
• Fix Version/s: 1.18.0, 1.16.3, 1.17.2
• Component/s: Runtime / Coordination
• Labels: pull-request-available
• Assignee: Weijie Guo
• Reporter: Weijie Guo
• Created: 05/Jun/23 13:04
• Updated: 07/Jun/23 09:01
• Resolved: 07/Jun/23 09:01

Description

We ran a job with slotmanager.number-of-slots.max = 10, taskmanager.numberOfTaskSlots = 10 and taskmanager.memory.process.size: 24000m. The resources of the cluster are sufficient, but no TaskManager can be allocated. It seems that there is a problem with the calculation logic of SlotManagerConfiguration#getMaxTotalMem. Due to the rounding down of division, the calculated MemorySize is too small.

Issue Links

• links to: GitHub Pull Request #22714