System running Oracle Linux 6.8 with UEK 4 4.1.12 and Sun ZFS Storage 7330 experiences Data validation i/o errors during takeover and failback testing (Doc ID 2205192.1)

Last updated on MARCH 23, 2023

Applies to:

Symptoms

 On a system running Oracle Linux Server release 6.8 with UEK4 4.1.12-37.5.1.el6uek.x86_64 and Sun ZFS Storage 7330 with Appliance Version: 2013.06.05.6.0,1-1.16, Data Validation errors occurred during takeover and failback.

When testing cached I/O (not direct IO) takeover/failback operations with
iSCSI single path LUNs (no multipath), vdbench will report block corruption:

10:31:18.693 localhost-0: ===> Data Validation Key miscompare.
10:31:18.693 localhost-0: ===> Data miscompare.

The Oracle Linux client /var/log/messages logs errors similar to:

Jul 8 09:59:07 localhost kernel: [166146.523386] Buffer I/O error on
dev sdc, logical block 11495902, lost sync page write

Ending in:

Jul 8 09:59:07 localhost kernel: [166146.587246] sd 26:0:0:0: timing
out command, waited 180s

And further confirmation that the data did not make it to the ext3
filesystem / LUN is at:

Jul 8 09:59:07 localhost kernel: [166146.626084] JBD: Detected IO
errors while flushing file data on sdc

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.