|
From: Christian F. <Chr...@t-...> - 2014-06-14 16:25:12
|
Andrew Engelbrecht wrote: > when i run long tests in debian sid, they usually get interrupted: > >> SMART Self-test log structure revision number 1 >> Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error >> # 1 Extended offline Completed without error 00% 77 - >> # 2 Short offline Completed without error 00% 74 - >> # 3 Extended captive Interrupted (host reset) 90% 73 - >> # 4 Extended captive Interrupted (host reset) 90% 3 - >> # 5 Extended offline Interrupted (host reset) 40% 3 - >> # 6 Conveyance offline Completed without error 00% 0 - >> # 7 Short offline Completed without error 00% 0 - > dmesg shows the following line near the interruption time for test #2: > >> [153384.901218] usb 1-1.1: reset high-speed USB device number 6 using ehci-pci > in the one successful long test, i didn't use the "-C" option. in the > prior test, which was interrupted, i did. for the other interrupted > tests, i tested it with and without the "-C" option while using an older > kernel. (perhaps i had "ejected" the disk through the gui before > starting... i don't actually know, since it was a long time ago.) Running captive tests via SAT may not work because smartctl does not pass a long command timeout to the SCSI pass-through ioctl. And even then it would require that the SAT layer in the USB bridge firmware supports such long (several hours) timeouts. Non-captive tests may be interrupted by a standby command sent to the drive after some time of I/O inactivity. Many USB bridges do this. Try to run some script which accesses the drive during the test. This may work or not: while true; do dd if=/dev/sdX iflag=direct count=1 of=/dev/null sleep 60 done Thanks, Christian |