Query with Large IN(...) Clause (Many Values) in Cluster With Many LDM Threads Fails After 6 Minutes: "ERROR 1296 (HY000): Got error 4008 'Receive from NDB failed' from NDBCLUSTER"

(Doc ID 2384318.1)

Last updated on APRIL 09, 2018

Applies to:

MySQL Cluster - Version 7.5 and later
Information in this document applies to any platform.

Symptoms

If you have a query with several hundred elements in an IN(...) clause with engine condition pushdown enabled and many LDM threads/partitions, then the query fails when the NDB API timeout of six minutes is encountered:

where the IN(...) clause contains 781 values and MaxNoOfExecutionThreads = 64 (which gives 32 LDM threads). The primary key of t1 is (t1_id, t2_id) and of t2 it is (t2_id).

Changes

The issue may start after increasing the number of LDM threads and rebuilding the tables.

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