DGenerateKane
Explorer
- Joined
- Sep 4, 2014
- Messages
- 95
I just got a used rackmount server off ebay last week (specs in sig), and I'm currently testing the first four of the eight drives I'm putting in it. The last four only just arrived today, otherwise they'd be in it getting tested as well. I followed Fester's Guide (https://www.familybrown.org/dokuwiki/doku.php?id=fester:hvalid_hdd) for hard drives tests step by step, and I'm a bit alarmed on the number of bad blocks on three of the drives during the destructive bad block test. I'm still running the tests so I can't post any results yet, just what I can see in the tmux sessions.
ada0 is the only one I'm not concerned with yet, as it doesn't have any bad blocks. Yet.
ada1 I thought was fine as it finished the bad blocks test itself, and was at the "reading and comparing" stage of the test. The next time I looked it was just outputting a ton of numbers in seemingly sequential order (Too fast for me to tell for sure) which I had already determined from looking at ada3 and ada4 that it was listing bad blocks, dozens per second. ada3 and ada4 already did that, and eventually aborted due to too many errors. I restarted the tests on both ada3 and ada4 and so far they are error free, and are much further along in the test then they got before spewing bad blocks everywhere. ada1 just aborted it's first attempt at a badblock test. I'm not sure if I should retry the test, or wait for the other three to finish theirs before trying something else, like checking the SMART data. At the rate they are going and assuming they don't abort due to too many bad blocks, I estimate at least 20 more hours. ada0 is 30 hours in, and is only 50% of the way through the reading and comparing stage. I would like to shut it down and put the last four drives in so I can test them as well, as I'm not sure if hot swapping is enabled in the BIOS. Stupid of me not to check that ahead of time.
This is the last bit visible in the tmux session for ada1:
I scrolled up as far as I could and it is just a crap ton more blocks listed as bad. If these drives are bad, why have two of them gone a lot further in the test with 0 errors so far this time? Oh, for reference the specific command I used was badblocks -b 4096 -vws /dev/daX for the destructive bad blocks test, in case it matters. Is there a different test I should do? Or are those three drives really bad and need to be RMA'd?
ada0 is the only one I'm not concerned with yet, as it doesn't have any bad blocks. Yet.
ada1 I thought was fine as it finished the bad blocks test itself, and was at the "reading and comparing" stage of the test. The next time I looked it was just outputting a ton of numbers in seemingly sequential order (Too fast for me to tell for sure) which I had already determined from looking at ada3 and ada4 that it was listing bad blocks, dozens per second. ada3 and ada4 already did that, and eventually aborted due to too many errors. I restarted the tests on both ada3 and ada4 and so far they are error free, and are much further along in the test then they got before spewing bad blocks everywhere. ada1 just aborted it's first attempt at a badblock test. I'm not sure if I should retry the test, or wait for the other three to finish theirs before trying something else, like checking the SMART data. At the rate they are going and assuming they don't abort due to too many bad blocks, I estimate at least 20 more hours. ada0 is 30 hours in, and is only 50% of the way through the reading and comparing stage. I would like to shut it down and put the last four drives in so I can test them as well, as I'm not sure if hot swapping is enabled in the BIOS. Stupid of me not to check that ahead of time.
This is the last bit visible in the tmux session for ada1:
Code:
1198666168 1198666169 1198666170 1198666171 1198666172 1198666173 1198666174 Too many bad blocks, aborting test done Testing with pattern 0x55: Too many bad blocks, aborting test073741823/0/0 error done Reading and comparing: Too many bad blocks, aborting test073741823/0/0 errors) done Testing with pattern 0xff: Too many bad blocks, aborting test073741823/0/0 error done Reading and comparing: Too many bad blocks, aborting test073741823/0/0 errors) done Testing with pattern 0x00: Too many bad blocks, aborting test073741823/0/0 error done Reading and comparing: Too many bad blocks, aborting test073741823/0/0 errors) done Pass completed, 1073741823 bad blocks found. (1073741823/0/0 errors)
I scrolled up as far as I could and it is just a crap ton more blocks listed as bad. If these drives are bad, why have two of them gone a lot further in the test with 0 errors so far this time? Oh, for reference the specific command I used was badblocks -b 4096 -vws /dev/daX for the destructive bad blocks test, in case it matters. Is there a different test I should do? Or are those three drives really bad and need to be RMA'd?