ZFS Scrub taking wildly inconsistent amounts of time to finish, no errors upon completion

Alright, getting the hardware stuff out of the way:
HP Z4 G4 OEM motherboard (Intel C422 Chipset)
Intel Xeon W-2223 CPU
64GB DDR4 ECC RDIMM (8x8)
Pool Hardware:
Western Digital WD120EFBX 12TB CMR drives x5 RAIDZ2
Samsung 980 Pro 2TB x2 Log VDEV (Mirrored)
Samsung 980 Pro 2TB x2 Cache

I have (more accurately, had) my ZFS scrubs set to occur weekly on Monday mornings at 0001, with my extended SMART weekly on Saturday mornings at 0001 so as to avoid both of them occurring at the same time. The issue that I’m having is that the scan that ran two weeks ago took about 5 days to complete and the one that ran most recently took 25 hours. Both of them completed without errors and without any repaired bytes, according to zfs status -v. SMART tests are also showing everything as good as well. I read somewhere that the scrub task takes time based on the storage usage of the pool, to that end my pool usage is 58.1%.

Anyone have any insight as to what could be going on here?

Any chance there was heavy usage during the smart test causing it to get pushed back & then both smart & a scrub ran at the same time?

1 Like