G Speed Posted May 4, 2022 Share Posted May 4, 2022 (edited) So I'm testing a new server.... Using Version: 6.10.0-rc5 Here is what's happening... Drives Spun Down... I click check parity, Drives spin up. I get 50-60MB/s per drive I Cancel... I spin drives down.. then I spin drives up Start parity, I get 200MB/s per drive Cancel.. Spin Down.. Click Parity... 50-60MB/s I can replicate this every time Log SHOWS nothing May 3 21:24:47 Tower2 emhttpd: read SMART /dev/sdb May 3 21:24:48 Tower2 emhttpd: spinning up /dev/sdh May 3 21:24:52 Tower2 emhttpd: read SMART /dev/sdh May 3 21:24:52 Tower2 emhttpd: read SMART /dev/sdd May 3 21:24:53 Tower2 emhttpd: spinning up /dev/sdi May 3 21:24:57 Tower2 emhttpd: read SMART /dev/sdi May 3 21:24:58 Tower2 emhttpd: read SMART /dev/sde May 3 21:24:58 Tower2 emhttpd: spinning up /dev/sdc May 3 21:25:10 Tower2 emhttpd: read SMART /dev/sdc May 3 21:25:10 Tower2 emhttpd: read SMART /dev/sdf May 3 21:25:10 Tower2 emhttpd: read SMART /dev/sdg May 3 21:25:11 Tower2 emhttpd: read SMART /dev/sdh May 3 21:25:11 Tower2 emhttpd: read SMART /dev/sdi May 3 21:25:11 Tower2 emhttpd: read SMART /dev/sdc May 3 21:25:20 Tower2 kernel: mdcmd (46): check May 3 21:25:20 Tower2 kernel: md: recovery thread: check P Q ... May 3 21:25:27 Tower2 kernel: mdcmd (47): nocheck Cancel May 3 21:25:27 Tower2 kernel: md: recovery thread: exit status: -4 May 3 21:25:38 Tower2 ntpd[1526]: kernel reports TIME_ERROR: 0x41: Clock Unsynchronized May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sdh May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sdg May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sdd May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sde May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sdb May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sdf May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sdc May 3 21:26:30 Tower2 emhttpd: spinning down /dev/sdi It's a Dell R530, with an H730 set to HBA <- Yes it's a proper IT Mode, full disk pass to OS, Smart Etc... This is when unraid calls for a spinup for parity check I will cancel the above.. SPIN DOWN the array.. then SPIN UP.. Start parity check.. Edited May 4, 2022 by G Speed Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 Downgraded to 6.9.2 same issue Quote Link to comment
Squid Posted May 4, 2022 Share Posted May 4, 2022 Have you let the parity check run for a couple of minutes (The snip shows it ran for 7 seconds)? What you're seeing is basically normal and the speed should be more or less identical after it all levels out and calms down. Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 1 hour ago, Squid said: Have you let the parity check run for a couple of minutes (The snip shows it ran for 7 seconds)? What you're seeing is basically normal and the speed should be more or less identical after it all levels out and calms down. I can try sure, but I have never seen this behavior on my other unraid systems Quote Link to comment
JorgeB Posted May 4, 2022 Share Posted May 4, 2022 IIRC this happened before to someone else, i.e., disks being slower if woke by activity, don't remember if you found if the problem were the disks or the controller. Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 (edited) 6 minutes ago, JorgeB said: IIRC this happened before to someone else, i.e., disks being slower if woke by activity, don't remember if you found if the problem were the disks or the controller. This is my concern..... Parity whatever.. I can spin up then start. But what's odd is when parity is running, and I pause it.. and wait.. All drives are spun up, if I resume.. same slow speed. If I spin down, then up again full speed! Edited May 4, 2022 by G Speed Quote Link to comment
JorgeB Posted May 4, 2022 Share Posted May 4, 2022 Try testing with the diskspeed docker, it might help identify the issue if it's just one disk/model, unless they are all the same. Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 7 minutes ago, JorgeB said: Try testing with the diskspeed docker, it might help identify the issue if it's just one disk/model, unless they are all the same. All perfect, did it yesterday... Quote Link to comment
JorgeB Posted May 4, 2022 Share Posted May 4, 2022 That's with the disks spun down before the test? Quote Link to comment
Squid Posted May 4, 2022 Share Posted May 4, 2022 23 minutes ago, G Speed said: I can try sure, but I have never seen this behavior on my other unraid systems I have always seen this, on every server and every OS version/ For the first min it's slow and then gets up to speed. Doesn't bother me because it always picks up to the max and it's the average time and speed when it's done that's important. Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 (edited) 6 minutes ago, Squid said: I have always seen this, on every server and every OS version/ For the first min it's slow and then gets up to speed. Doesn't bother me because it always picks up to the max and it's the average time and speed when it's done that's important. First minute sure, not 16mins 26mins still same Stop parity, try again same Spin down, spin up Edited May 4, 2022 by G Speed Quote Link to comment
trurl Posted May 4, 2022 Share Posted May 4, 2022 Why did you give us a screenshot showing it has run less than a minute? Demonstrate it is still happening after things settle. Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 (edited) 10 minutes ago, trurl said: Why did you give us a screenshot showing it has run less than a minute? Demonstrate it is still happening after things settle. Total size:6 TB Elapsed time:44 minutes Current position:525 GB (8.8 %) Estimated speed:193.1 MB/sec Estimated finish:7 hours, 53 minutes Sync errors corrected: 0 Everything is 100% fine when I spin up drives Edited May 4, 2022 by G Speed Quote Link to comment
trurl Posted May 4, 2022 Share Posted May 4, 2022 2 minutes ago, G Speed said: Total size:6 TB Elapsed time:44 minutes Current position:525 GB (8.8 %) Estimated speed:193.1 MB/sec Estimated finish:7 hours, 53 minutes Sync errors corrected: 0 That looks excellent (of course). Can you show us the problem after it has run for a while? Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 (edited) 5 minutes ago, trurl said: That looks excellent (of course). Can you show us the problem after it has run for a while? I think you missed the original issue, the above is me spinning up the drives and starting.. it will do the parity scan at full speed "according to the benchmarks" for the full duration. On the other hand When unraid starts parity, when disk are spundown It won't hit over 60MB/s, 1min..40mins.. doesn't matter.. doesn't catch up.. doesn't slow down Edited May 4, 2022 by G Speed Quote Link to comment
trurl Posted May 4, 2022 Share Posted May 4, 2022 I understand your description just fine. The (tiny) screenshot in an earlier post seemed to show it being slow after 6 minutes, but the description in that screenshot also talks about pausing. Can you give a screenshot that shows it being slow after it has run continuously for several minutes? Quote Link to comment
G Speed Posted May 4, 2022 Author Share Posted May 4, 2022 1 hour ago, G Speed said: All perfect, did it yesterday... 5 minutes ago, trurl said: I understand your description just fine. The (tiny) screenshot in an earlier post seemed to show it being slow after 6 minutes, but the description in that screenshot also talks about pausing. Can you give a screenshot that shows it being slow after it has run continuously for several minutes? Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.