Excessive Watchdog Warnings During Memory Compaction/Defragmentation: "[ndbd] WARNING -- Ndb kernel thread ... is stuck in: Performing Send elapsed=..."; "Forced node shutdown completed. Caused by error 6050: 'WatchDog terminate, ..." (Doc ID 2161913.1)

Last updated on MARCH 08, 2017

Applies to:

MySQL Cluster - Version 7.2 and later
Generic Linux

Symptoms

One or more data nodes suffers for an excessive amount of watchdog warnings for a kernel thread being stuck in Performing Send, for example in the data node's out log:

The numbers are the number of free blocks per size with the larger blocks to the right. So the above shows that there are none of the largest block sizes free.

If Transparent Huge Pages (THP) is enabled, the issue may occur with less memory fragmentation.



Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms