My Oracle Support Banner

Manual Server and Service Migration Fails to Succeed When Going Back to Preferred Server (Doc ID 1596057.1)

Last updated on NOVEMBER 22, 2022

Applies to:

Oracle WebLogic Server - Version 10.3.4 and later
Information in this document applies to any platform.


When a manual service migration is started for one of the servers this goes fine. After this, the physical box gets bounced and then the managed server gets restarted as expected. At this point, the primary cluster does not renew the lease of it, and seems that it does not recognize it as a valid member of it. For such reason whenever it is tried to migrate the server back to its preferred candidate, even though the console says that it was successfully migrated, non of the targets are running on the newly migrated server.


Consider the following as an example architecture:

Managedserver_1 ----Machine1
Managedserver_2 ----Machine2
Managedserver_3 ----Machine3
Managedserver_4 ----Machine4
weblogic_admin_1 --- Machine5

1. Testing service migration:

a) Bring up the managed server's one by one. The server that is started first will host all the migratable targets. Once all the managed servers are up and running, move the migratable targets to their corresponding preferred servers.
b) Shutdown one of the managed server, say Managedserver_2.
c) Observe the migratable targets Monitoring page and view the "Current hosting server" for Managedserver_2 (migratable), let's say, its migrated to Managedserver_4
d) Start Managedserver_2.
e) Manually migrate Managedserver_2 (migratable) back to Managedserver_2

This scenario works fine.

2. Testing whole server migration:

a) Once all the managed servers are up and running. Shutdown one of the machine's (machine that is not Cluster ), say Machine2.
b) Observe in WLS Admin console-->Domain_name-->Monitoring-->Migration, the machine where Managedserver_2 is migrated to. Lets assume, its migrated to Machine4.
c) Observe in migratable targets Monitoring page, Managedserver_2 (migratable) current hosting server. Lets assume its migrated to Managedserver_4.
d)Reboot Machine2.
e) Start the Nodemanager (if its not already started as background process)
f) Shutdown "Managedserver_2" from WLS Admin Console and restart it again.
g) Observe the "current hosting machine" for Managedserver_2 after its restarted. It should point to Machine2 which indicates a successful Whole server migration and failback.

The above scenario works fine.

3. Testing both whole server and service migration:

a) Follow steps (a) through (g) in the Test Scenario 2.
b) Manually migrate Managedserver_2 (migratable) from Managedserver_4 to Managedserver_2.
c) Console shows the status as "succeeded". However, there will not be any indication in WLS logs (with migration debugs enabled) on this manual migration.
d) Manually migrate Managedserver_3 (migratable) from Managedserver_3 to Managedserver_2. Same thing, console says it "succeeded", but the migration did not happen at all.
e) Any manual migration to Managedserver_2 does not happen (after a successful whole server migration and a failback)

This scenario does not work.

When this is the case, then the following will appear after enabling Singleton Services Debug:



To view full details, sign in with your My Oracle Support account.

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

In this Document

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