[PCA 2.x] Virtual Machines are Moved Automatically Between Compute Nodes while DRS (Distributed Resource Scheduler) is not Enabled (Doc ID 2297653.1)

Last updated on OCTOBER 02, 2022

Applies to:

Symptoms

Guests are moved automatically between Compute Nodes, even though DRS (Distributed Resource Scheduler) is not enabled.

The AdminServer.log shows the following:

<2017-08-16T14:58:30.375+0200> <Info> <com.oracle.ovm.mgr.task.loadbalancer.PoolLoadBalancer> <xxxxx> <AdminServer> <TaskThread_4c:4c:45:44:00:56:46:10:80:31:b1:c0:4f:39:35:32_EVACUATE-146> <> <> <7e1c4e2a-77f8-4977-be91-b30541c0b715-00000011> <1502888310375> <BEA-000000> <THREAD STARTED: Server evacuate called for server 4c:4c:45:44:00:56:46:10:80:31:b1:c0:4f:39:35:32>
####<2017-08-16T14:58:30.441+0200> <Info> <com.oracle.ovm.mgr.event.ovs.OvsEventProcessorTask> <xxxxx> <AdminServer> <EventProcessor-5> <> <> <7e1c4e2a-77f8-4977-be91-b30541c0b715-00000011> <1502888310441> <BEA-000000> <370 server notification events were processed in the last hour>
####<2017-08-16T14:58:30.540+0200> <Info> <com.oracle.ovm.mgr.task.loadbalancer.PoolLoadBalancer> <xxxxx> <AdminServer> <TaskThread_4c:4c:45:44:00:56:46:10:80:31:b1:c0:4f:39:35:32_EVACUATE-146> <> <> <7e1c4e2a-77f8-4977-be91-b30541c0b715-00000011> <1502888310540> <BEA-000000> <Moving VM testvm from server yyyyy to server zzzzz in job: 1502888310529>

The reference to <com.oracle.ovm.mgr.task.loadbalancer.PoolLoadBalancer> would imply this is related to DRS.

Changes

Compute Node where the Guests are running is put in maintenance mode.

This can be seen in the AdminServer.log:

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.