[HDFS-14981] BlockStateChange logging is exceedingly verbose - ASF JIRA

Voters

Watch issue

Watchers

Create sub-task

Link

Clone

Update Comment Author

Replace String in Comment

Update Comment Visibility

Delete Comments

XML

Word

Printable

JSON

Details

Type: Bug
Status: Resolved
Priority: Major
Resolution: Duplicate
Affects Version/s: None
Fix Version/s: None
Component/s: logging
Labels:
None

Description

On a moderately loaded cluster, name node logs are flooded with entries of INFO BlockStateChange..., to the tune of ~30 lines per millisecond. This provides operators with little to no usable information. I suggest reducing this log message to DEBUG. Perhaps this information (and other logging related to it) should be directed to a dedicated block-audit.log file that can be queried, rotated on a separate schedule from the log of the main process.