Oracle Database 19.7 Automatic Startup on OLS 8.3

Radu N.

Good morning all.

I'm stuck with an issue on Oracle Database startup. The DB version and OS version are in the Question title.

The issue I'm facing is that sometimes, but not always, on server reboot, the Oracle Service doesn't start.

The server's disks are mounted like this:

/dev/mapper/ARCLOG-ARCLOG_VOL   350G 3.0G 347G  1% /oraarchive

/dev/mapper/BACKUP-BACKUP_VOL   1.4T 9.8G 1.4T  1% /orabackup

/dev/mapper/ORADATA-ORADATA_VOL 1.5T 852G 684G 56% /u01

The error I got in /var/log/messages is the following one:

Mar 16 10:42:18 trn-oracle systemd[1241]: ora_db.service: Failed to execute command: No such file or directory

Mar 16 10:42:18 trn-oracle systemd[1241]: ora_db.service: Failed at step EXEC spawning /u01/app/oracle/product/19.0.0/dbhome_1/bin/dbstart: No such file or directory

Mar 16 10:42:18 trn-oracle systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.

Mar 16 10:42:18 trn-oracle systemd[1]: ora_db.service: Control process exited, code=exited status=203

Mar 16 10:42:18 trn-oracle systemd[1]: ora_db.service: Failed with result 'exit-code'.

That happens because the mount of /u01 partition chronologically finishes AFTER the start or Oracle Service.

My service file is:

------------------------------------------------------------------------------------------------------------------------

Description=The Oracle 19.7 Database Service

After=rsyslog.target network.target local-fs.target

#Requires=network.target u01.mount orabackup.mount

#RequiresMountsFor=/u01

#RequiresMountsFor=/orabackup

#ConditionPathIsReadWrite=/u01

#ConditionPathIsReadWrite=/orabackup

[Service]

Type=forking

# Systemd ignores PAM limits, so set any necessary limits in the service.

# Not really a bug, but a feature. https://bugzilla.redhat.com/show_bug.cgi?id=754285

LimitMEMLOCK=infinity

LimitNOFILE=65535

EnvironmentFile=/etc/sysconfig/env.ora-trn

# Type=simple

# idle: similar to simple, the actual execution of the service binary is delayed

# until all jobs are finished, which avoids mixing the status output with shell output of services.

# RemainAfterExit=yes

User=oracle

Group=oinstall

#Restart=no

ExecStart=/u01/app/oracle/product/19.0.0/dbhome_1/bin/dbstart /u01/app/oracle/product/19.0.0/dbhome_1

ExecStop=/u01/app/oracle/product/19.0.0/dbhome_1/bin/dbshut /u01/app/oracle/product/19.0.0/dbhome_1

[Install]

WantedBy=multi-user.target

------------------------------------------------------------------------------------------------------------------------

I tried some options in the "After=" condition (the ones commented below) without success.

Any ideas are welcome because I'm stuck here for about two days.

Thank you very much for your time.

Regards.

Radu N.

Nik

Hi.

According https://unix.stackexchange.com/questions/246935/set-systemd-service-to-execute-after-fstab-mount

1. I do not see [Unit] section at provided service file.
2. check systemctl list-units | grep mount

You should find services like: u01.mount orabackup.mount oraarchive.mount

Add this service to section Unit as Required and After

Regards,

Nik

Radu N.

Hi.

TY for answering.

I already saw that link, and I already test all the options below. No luck.

#Requires=network.target u01.mount orabackup.mount

#RequiresMountsFor=/u01

#RequiresMountsFor=/orabackup

#ConditionPathIsReadWrite=/u01

#ConditionPathIsReadWrite=/orabackup

Nik

Hi.

At this moment this option commented, so not work.

I do not see options After.

You can stop Oracle service, stop u01.mount and try start Oracle service back and check logs.