dNFS is kicking my ass
From: DOUG KUSHNER <dougk5_at_cox.net>
Date: Mon, 23 Oct 2023 12:30:27 -0700 (MST)
Message-ID: <2041141621.1919206.1698089427356_at_myemail.cox.net>
Most of my time is spent on Exadata, but have a need to set up 19.20 on OL8.8 with OFA on file systems mounted from a Dell Unity. We have followed Dell's Oracle Best Practices guide, Oracle's documentation and everything that I could Google on the subject.
Date: Mon, 23 Oct 2023 12:30:27 -0700 (MST)
Message-ID: <2041141621.1919206.1698089427356_at_myemail.cox.net>
Most of my time is spent on Exadata, but have a need to set up 19.20 on OL8.8 with OFA on file systems mounted from a Dell Unity. We have followed Dell's Oracle Best Practices guide, Oracle's documentation and everything that I could Google on the subject.
After 2 days of working with Oracle Support, their conclusion is that my setup is ok and are questioning the storage setup. There don't appear to be any known dNFS bugs that haven't been fixed in 19.20.
There are 2 paths to the storage from the server and each path is mountable with kNFS, but not with dNFS. The control channel has not been assigned to a separate interface. The database was created with dbca over kNFS.
mount points are /u02 for oradata and /u01/app/oracle/fast_recovery_area for the FRA.
Can anyone share some troubleshooting tips? I set events in a pfile and attempted to mount the database, but the resulting debug info was inconclusive.
event="10298 trace name context forever, level 1" event="19392 trace name context forever, level 8" event="19394 trace name context forever, level 8" event="19396 trace name context forever, level 2" event="19398 trace name context forever, level 128"
Regards,
Doug
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Oct 23 2023 - 21:30:27 CEST