Adding a PDB (and Service) to a RAC database -- service running on only 1 Instance

On my existing, RAC database I have :

[oracle@node1 ~]$ srvctl status database -db DB21CRAC
Instance DB21CRAC1 is running on node node1
Instance DB21CRAC2 is running on node node2
[oracle@node1 ~]$
[oracle@node1 ~]$ srvctl status service -d DB21CRAC -s hemantpdb
PRKO-2017 : Service hemantpdb does not exist for database DB21CRAC.
[oracle@node1 ~]$ srvctl status service -db DB21CRAC -service hemantpdb -pdb HEMANTPDB
PRKO-2017 : Service hemantpdb does not exist for database DB21CRAC.
[oracle@node1 ~]$
[grid@node1 ~]$ lsnrctl status LISTENER_SCAN1

LSNRCTL for Linux: Version 21.0.0.0.0 - Production on 13-FEB-2024 15:06:11

Copyright (c) 1991, 2021, Oracle.  All rights reserved.

Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=IPC)(KEY=LISTENER_SCAN1)))
STATUS of the LISTENER
------------------------
Alias                     LISTENER_SCAN1
Version                   TNSLSNR for Linux: Version 21.0.0.0.0 - Production
Start Date                13-FEB-2024 14:51:13
Uptime                    0 days 0 hr. 14 min. 57 sec
Trace Level               off
Security                  ON: Local OS Authentication
SNMP                      OFF
Listener Parameter File   /u01/app/21.3.0.0/grid/network/admin/listener.ora
Listener Log File         /u01/app/grid/diag/tnslsnr/node1/listener_scan1/alert/log.xml
Listening Endpoints Summary...
  (DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=LISTENER_SCAN1)))
  (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=192.168.56.115)(PORT=1521)))
Services Summary...
Service "0f488ad896262f80e0636f38a8c0fc18" has 2 instance(s).
  Instance "DB21CRAC1", status READY, has 1 handler(s) for this service...
  Instance "DB21CRAC2", status READY, has 1 handler(s) for this service...
Service "DB21CRAC" has 2 instance(s).
  Instance "DB21CRAC1", status READY, has 1 handler(s) for this service...
  Instance "DB21CRAC2", status READY, has 1 handler(s) for this service...
Service "DB21CRACXDB" has 2 instance(s).
  Instance "DB21CRAC1", status READY, has 1 handler(s) for this service...
  Instance "DB21CRAC2", status READY, has 1 handler(s) for this service...
Service "hemantpdb" has 2 instance(s).
  Instance "DB21CRAC1", status READY, has 1 handler(s) for this service...
  Instance "DB21CRAC2", status READY, has 1 handler(s) for this service...
The command completed successfully
[grid@node1 ~]$
[oracle@node1 ~]$SQL> select inst_id, name, network_name from gv$services order by 1;

   INST_ID NAME             NETWORK_NAME
---------- ---------------- ----------------
         1 hemantpdb        hemantpdb
         1 DB21CRAC         DB21CRAC
         1 SYS$BACKGROUND
         1 DB21CRACXDB      DB21CRACXDB
         1 SYS$USERS
         2 hemantpdb        hemantpdb
         2 DB21CRAC         DB21CRAC
         2 SYS$BACKGROUND
         2 DB21CRACXDB      DB21CRACXDB
         2 SYS$USERS

10 rows selected.

SQL>
SQL> select inst_id, con_id, name, open_mode from gv$pdbs order by 1,2;

   INST_ID     CON_ID NAME             OPEN_MODE
---------- ---------- ---------------- ----------
         1          2 PDB$SEED         READ ONLY
         1          3 HEMANTPDB        READ WRITE
         2          2 PDB$SEED         READ ONLY
         2          3 HEMANTPDB        READ WRITE

SQL>

Which means that I have created a custom PDB called "HEMANTPDB" and there is a default service called "hemantpdb" on each instance.  However, this services is NOT listed when I check via srvctl. 

This is because srvctl queries the cluster for information about services.  Service"hemantpdb" is created by default when I run CREATE PLUGGABLE DATABASE HEMANTPDB.

The "add service" command to srvctl has a large number of parameters.

So, I can add a new service and configure Transparent Application Failover for SELECT failover but with only the first instance(DB21CRAC1) as the only one to start the service on initially{and the second instance (DB21CRAC2) as the alternate} :

[oracle@node1 ~]$ srvctl add service -db DB21CRAC  -service newservice -preferred DB21CRAC1 -available DB21CRAC2 -tafpolicy BASIC -failovertype SELECT -pdb HEMANTPDB
[oracle@node1 ~]$
[oracle@node1 ~]$ srvctl config service -db DB21CRAC  -service newservice
Service name: newservice
Server pool:
Cardinality: 1
Service role: PRIMARY
Management policy: AUTOMATIC
DTP transaction: false
AQ HA notifications: false
Global: false
Commit Outcome: false
Reset State: NONE
Failover type: SELECT
Failover method:
Failover retries:
Failover delay:
Failover restore: NONE
Connection Load Balancing Goal: LONG
Runtime Load Balancing Goal: NONE
TAF policy specification: BASIC
Edition:
Pluggable database name: HEMANTPDB
Hub service:
Maximum lag time: ANY
SQL Translation Profile:
Retention: 86400 seconds
Failback :  no
Replay Initiation Time: 300 seconds
Drain timeout:
Stop option:
Session State Consistency: DYNAMIC
GSM Flags: 0
Service is enabled
Preferred instances: DB21CRAC1
Available instances: DB21CRAC2
CSS critical: no
[oracle@node1 ~]$
[oracle@node1 ~]$ srvctl start service -db DB21CRAC -service newservice
[oracle@node1 ~]$ srvctl status service -db DB21CRAC -service newservice
Service newservice is running on instance(s) DB21CRAC1
[oracle@node1 ~]$

If the database instance DB21CRAC1 fails (crashes), the ClusterWare starts the service on DB21CRAC2.  Here I kill the DB21CRAC process and then verify that the service has restarted on DB1CRAC2  :

[oracle@node1 ~]$ srvctl status service -db DB21CRAC -service newservice
Service newservice is running on instance(s) DB21CRAC1
[oracle@node1 ~]$ ps -ef |grep smon
oracle    2951     1  0 16:05 ?        00:00:00 ora_smon_DB21CRAC1
root      3521     1  1 14:50 ?        00:00:48 /u01/app/21.3.0.0/grid/bin/osysmond.bin
grid      4068     1  0 14:50 ?        00:00:00 asm_smon_+ASM1
oracle    4146 25526  0 16:05 pts/0    00:00:00 grep --color=auto smon
[oracle@node1 ~]$ kill -9 2951
[oracle@node1 ~]$ srvctl status service -db DB21CRAC -service newservice
Service newservice is running on instance(s) DB21CRAC2
[oracle@node1 ~]$

In this case, connections using this service name ("newservice") will connect (failover) to the 2nd database instance running on node2 of the Cluster.