Comment 14 for bug 1271268

I added my address to the CC: list because something similar happens to me. I
have two 4GB seagate USB3 "desktop expansion drives" which change device (say
from sdc to sdd) if they are being written to _very_ slowly (an image copy of
my mostly zeroed Win8 partition compressed by xz for instance). Otherwise they
behave perfectly well - never as much as a burp. My approach to my problem is
"don't do that then" so I have not bothered to report the bug (and like I said
my bug is only _slightly_ similar -- apart from complaining that the drive it
was writing to has "disappeared" the kernel does not log any other errors.) My
drives never "disconnect" when not not being used, or when read or written to
at "normal" (very fast!) speeds. (Just my 2 cents worth -- I didn't want to
distract attention from this rather different bug by mentioning this before.)