I faced with that error during 19c standalone installation on linux.
In Grid Installation step I get below error:
INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: Configuring HA resource failed. INFO: [Nov 17, 2020 12:05:21 PM] PRCR-1079 : Failed to start resource ora.asm INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: PRCR-1079 : Failed to start resource ora.asm INFO: [Nov 17, 2020 12:05:21 PM] CRS-2674: Start of 'ora.LISTENER.lsnr' on 'veridatait' failed INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: CRS-2674: Start of 'ora.LISTENER.lsnr' on 'nodename' failed INFO: [Nov 17, 2020 12:05:21 PM] CRS-5017: The resource action "ora.asm start" encountered the following error: INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: CRS-5017: The resource action "ora.asm start" encountered the following error: INFO: [Nov 17, 2020 12:05:21 PM] ORA-00845: MEMORY_TARGET not supported on this system INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: ORA-00845: MEMORY_TARGET not supported on this system INFO: [Nov 17, 2020 12:05:21 PM] . For details refer to "(:CLSN00107:)" in "/u01/app/oracle/diag/crs/nodename/crs/trace/ohasd_oraagent_oracle.trc". INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: . For details refer to "(:CLSN00107:)" in "/u01/app/oracle/diag/crs/nodename/crs/trace/ohasd_oraagent_oracle.trc". INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: INFO: [Nov 17, 2020 12:05:21 PM] CRS-2674: Start of 'ora.asm' on 'veridatait' failed INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: CRS-2674: Start of 'ora.asm' on 'nodename' failed INFO: [Nov 17, 2020 12:05:21 PM] Skipping line: INFO: [Nov 17, 2020 12:05:21 PM] Completed Plugin named: asmca
So The new ASM functionality uses /dev/shm on Linux for SGA and PGA management. The errors occur if either MEMORY_TARGET or MEMORY_MAX_TARGET is configured larger than the configured /dev/shm size
or if /dev/shm is mounted incorrectly.
So to can resolve it, login to root user and edit the fstab and increase the value of /dev/shm. Like tmpfs /dev/shm tmpfs size=100G .
Important point is You need to set the tmpfs size more than the required memory target.
- Edit /et/fstab file with correct size
[root@nodename ]$vi /etc/fstab
tmpfs /dev/shm tmpfs size=30g 0 0 ===>make size 100G
[root@nodename~]# umount /dev/shm [root@nodename~]# mount /dev/shm
[root@nodename]$ df -h /dev/shm
Filesystem Size Used Avail Use% Mounted on
tmpfs 100G 0 100G 0% /dev/shm
Finally after making changes started the database and it worked perfectly.