[HDFS-14657] Refine NameSystem lock usage during processing FBR

• Type: Improvement
• Status: Patch Available
• Priority: Major
• Resolution: Unresolved
• Assignee: Chen Zhang
• Reporter: Chen Zhang
• Created: 17/Jul/19 09:19
• Updated: 02/Dec/22 02:25

Description

The disk with 12TB capacity is very normal today, which means the FBR size is much larger than before, Namenode holds the NameSystemLock during processing block report for each storage, which might take quite a long time.

On our production environment, processing large FBR usually cause a longer RPC queue time, which impacts client latency, so we did some simple work on refining the lock usage, which improved the p99 latency significantly.

In our solution, BlockManager release the NameSystem write lock and request it again for every 5000 blocks(by default) during processing FBR, with the fair lock, all the RPC request can be processed before BlockManager re-acquire the write lock.

Issue Links

• is duplicated by: HDFS-14658 Refine NameSystem lock usage during processing FBR
• is duplicated by: HDFS-14659 Refine NameSystem lock usage during processing FBR
• is related to: HDFS-11313 Segmented Block Reports
• is related to: HADOOP-16452 Increase ipc.maximum.data.length default from 64MB to 128MB