might be a flaky test/operation, happened while building in i386 buster environment:

neurodebian@smaug ..7.20190730+git103-g070fbd693-1~ndall+1 % grep -B20 -A1 FAIL git-annex_7.20190730+git103-g070fbd693-1~ndall+1_i386.build
             20 100%    0.00kB/s    0:00:00 (xfr#1, to-chk=0/1)
(checksum...) (recording state in git...)
ok
(recording state in git...)
OK (1.00s)
    move (ssh remote):                                    move foo (from origin...) 
(checksum...) ok                  
(recording state in git...)
move foo (to origin...) 
ok                                
(recording state in git...)
OK (0.99s)
    copy:                                                 copy foo (from origin...) 
SHA256E-s20--e394a389d787383843decc5d3d99b6d184ffa5fddeec23b911f9ee7fc8b9ea77
             20 100%    0.00kB/s    0:00:00 (xfr#1, to-chk=0/1)

SHA256E-s20--e394a389d787383843decc5d3d99b6d184ffa5fddeec23b911f9ee7fc8b9ea77
             20 100%    0.00kB/s    0:00:00 (xfr#1, to-chk=0/1)
failed
copy: 1 failed
FAIL (0.34s)
      Test.hs:530:

I have tried to rebuild (debian/rules binary) second time while in the same environment -- didn't succeed... but then I proceeded to build amd64 fine, and redid i386 build just fine

fixed though w/o a true root cause analysis --Joey