https://slickdeals.net/f/15350722-costco-members-starts-11-15-seagate-backup-plus-hub-8tb-external-desktop-hard-drive-119-99?p=150687460#post150687460

Anyone know how to check if a drive is SMR kr CMR? I'm curious because I have several shucked drives that I use in Raid10. It's functioning just fine for over 3 yrs now but out of curiosity I want to check the drives.
It's hardware raid10 on dell h730 card

According to my research several days ago, no easy way to test.

If the manufacturer publishes a list of SMR drivers, and yours are on it, then definitely they are SMR. But the list is not exhaustive. If yours are not on the list, it doesn't mean it is CMR.

For testing, usually it involves carefully monitoring the writing speed while overloading the cache deliberately. Or, rewrite the same location of the drive and monitor the speed. But these tests may fail if TRIM is used or filesystem like ZFS is used. Anyway, tests are complicated and there is no guarantee for CMR.

Some also claim it can be determined from storage density listed in manufacturer specifications. But it was a great burden to me to check and decypher those information…

It seems many people agree high-end drivers with >10TB capacity are CMR; while those with [3.5in, >1TB per disk, usually 2T, 4T, 8T], or [2.5in, >0.5T per disk, usually 1T, 2T, 4T], are very much likely to be SMR.

Anyway, SMR is not devil, it can be quartered by a well backup manner. For me, I just want to save some hassle, so $200 for 12TB CMR is still an appleaing deal to me.

I at first got very confused about using smartctl or badblocks to test my brand new HDD, until I read this discussion of How are SMART selftests related to badblocks?.

• SMART should have automatically cleared all bad sectors by remapping, thus badblocks should not find anything if SMART says passed. So, do a SMART check first.
• If badblocks found something, it genearlly means not only a bad sector, but also that SMART is failing. For a new HDD, warranty should be asked. For an old one, replace it immediately.

Eventually I used this line for testing

badblocks -wsv -b 4096 /dev/device  # https://wiki.archlinux.org/title/badblocks#Read-write_test_(warning:_destructive)

-b 4096 is added to support 14T HDD.

badblocks read-write (destructive) testing is very slow.

Something I read:

• https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=554794
• https://davejansen.com/use-badblocks-to-check-larger-hard-drives/