Binlog corruption in Galera

Description

Seeing binlog corruption and other errors on slave nodes.

Slave bootstrap procedure:
Dump on cluster node 1:

On each slave node:

Sometimes this works sometimes it does not. If it works, it dies minutes later with errors like:

Binlog issues:

Slave errors:

or

or

more

Environment

6x Galera Nodes:
Kernel: 3.10.0-229.14.1.el7.x86_64
Disk: Local Virtio 200GB
Hypervisor: CentOS7 KVM
vCPUS: 4
Memory: 16GB

4x Slaves:
Kernel: 3.10.0-229.14.1.el7.x86_64
Disk: Local Virtio 40GB
Hypervisor: CentOS7 KVM
vCPUS: 2
Memory: 4GB

1x Garb:
Kernel: 3.10.0-229.14.1.el7.x86_64
Physical Node

MariaDB Packge: MariaDB-server-10.1.8-1.el7.centos.x86_64

Slave Config:

Galera Config:

Assignee

Nirbhay Choubey

Reporter

Bryan Seitz

Labels

Components

Sprint

None

Fix versions

Affects versions

Priority

Blocker
Configure