|
|
|
|
|
|
by mwint
980 days ago
|
|
|
If you’re serious - any profession in which you’re paid $x/hr, and a cheap knockoff cable could fail and burn 130/x hours of time. Extra worth it because Murphy’s law will make it fail at the worst possible moment. Or more likely, your cheap knockoff cable works well enough except when Joe from accounting is warming his Chef Boyardee in the microwave across the building. That kind of intermittent heisenbug can stealthily burn dozens of valuable hours and not show up on any accounting line. |
|
|
Plugged the thing in with two 16TB disks on a Linux machine, and proceeded to test the disks / enclosure to make sure disks were fine before moving any serious data on those - sometimes you get duds or they get damaged during transport - and that the setup was stable and performant enough for my needs.
Well, after a few TBs written it looked like one of the two disks was having write timeouts, operations hanged SMART counters were going up like crazy. It really looked like one of the disks had damage from transport. Tried it again just to be sure and now it was the second disk, with the first one being perfectly fine! More tests showed that the failure was fairly random.
So, buggy enclosure? Well, out of a link I tried the other cable (USB-C <-> USB-A) that came with the enclosure. That was slower, but writing was all fine.
Tried the original C-C and errors came up again. I had other C-C cables around of various but not spectacular quality, all of them were showing various issues, some even failing right on USB enumeration. Ultimately I decided to not cheap out and find a good cable, and I had an Apple TB3 cable - which does support USB 3.2 2x2 - for some reason that I didn't try as I wanted to keep it for something else. It's been rock solid and plenty fast.
My guess is that root cause was interference with other cables unable to cope at higher speeds.