MGMTDB undo tablespace
I have 2 pairs of RAC nodes (12.1.0.2.0) all patched to the same level on RHEL 6.5.
- The first pair of nodes is for 2 PROD database
- The second pair of nodes has the standby databases for PROD as well as 6 other test databases.
For the PROD nodes the size of the GI MGMTDB is static, but for the nodes containing the standby and test database the size of the UNDO tablespace for the MGMTDB database is steadily increasing.
i.e.
PROD MGMTDB UNDO tablespace - 126 Mb
Test MGMTDB UNDO tablespace - 15 Gb
I can see that datafiles on both PROD and test are autoextensible.