In trying to debug xfstest 129 I noticed odd behavior to Windows server test 129 does 4 steps, and the second one is the very slow step $here/src/looptest -i 100000 -r -w -b 8192 -s $SCRATCH_MNT/looptest/looptest1.tst $here/src/looptest -i 10000 -t -r -w -s -b 102400 $SCRATCH_MNT/looptest/looptest2.tst $here/src/looptest -i 50000 -r -w -b 256 -s $SCRATCH_MNT/looptest/looptest3.tst $here/src/looptest -i 2000 -o -r -w -b 8192 -s $SCRATCH_MNT/looptest/looptest4.tst So the problem is the second invocation of looptest - when it is doing 10,000 102,400 byte sequential writes and reads passing in the truncate flag. What I see in the network trace (at least to Windows server) that is troubling is lots of "STATUS_PENDING" errors in the write responses. It appears that the test runs much faster to Samba -- Thanks, Steve
