pyrater Posted March 17, 2017 Share Posted March 17, 2017 (edited) I dont think anything is wrong per say but god is it slow (2 days) to do a parity check. Does anyone have any tips to fix this / speed it up? System: M/B: Supermicro - H8DM8-2 CPU: Six-Core AMD Opteron™ 2419 EE @ 1800 X2 for 12 cores Cache: 768 kB, 3072 kB, 6144 kB Memory: 8 GB (max. installable capacity 8 GB) Network: eth0: 1000 Mb/s, full duplex, mtu 1500 Kernel: Linux 4.9.10-unRAID x86_64 Array size: 34 TB. Parity Drive: Total size:3 TB Elapsed time:20 hours, 21 minutes Current position:1.55 TB (51.7 %) Estimated speed:18.0 MB/sec Estimated finish:22 hours, 21 minutes Sync errors corrected:0 Log: Mar 15 16:12:18 Icarus kernel: mdcmd (304): spindown 1 Mar 15 16:12:21 Icarus kernel: mdcmd (305): spindown 2 Mar 15 16:12:22 Icarus kernel: mdcmd (306): spindown 3 Mar 15 16:12:25 Icarus kernel: mdcmd (307): spindown 4 Mar 15 16:12:26 Icarus kernel: mdcmd (308): spindown 5 Mar 15 17:09:10 Icarus kernel: mdcmd (309): spindown 11 Mar 15 17:29:34 Icarus kernel: mdcmd (310): spindown 1 Mar 15 17:29:44 Icarus kernel: mdcmd (311): spindown 3 Mar 15 17:29:46 Icarus kernel: mdcmd (312): spindown 5 Mar 15 17:49:37 Icarus kernel: Plex Media Scan[15506]: segfault at 2afdd05568ff ip 00002afd852d642c sp 00002afd8a3556a0 error 4 in libjemalloc.so.1[2afd8523b000+133000] Mar 15 17:50:30 Icarus kernel: Plex Media Scan[16159]: segfault at 2ace351568ff ip 00002acde9cb042c sp 00002acdef5fe700 error 4 in libjemalloc.so.1[2acde9c15000+133000] Mar 15 17:50:30 Icarus kernel: Plex Media Scan[16157]: segfault at 2ad56bf1e5c0 ip 00002acde9c49b08 sp 00007ffc9cc31b30 error 6 in libjemalloc.so.1[2acde9c15000+133000] Mar 15 18:52:51 Icarus kernel: mdcmd (313): spindown 10 Mar 15 18:54:36 Icarus kernel: mdcmd (314): spindown 2 Mar 15 18:54:39 Icarus kernel: mdcmd (315): spindown 4 Mar 15 18:54:42 Icarus kernel: mdcmd (316): spindown 1 Mar 15 18:54:42 Icarus kernel: mdcmd (317): spindown 3 Mar 15 19:02:14 Icarus kernel: mdcmd (318): spindown 9 Mar 15 19:12:24 Icarus kernel: Plex Media Scan[2975]: segfault at 0 ip (null) sp 00002b5b166bf8d8 error 14 in Plex Media Scanner[400000+42e000] Mar 15 20:17:08 Icarus kernel: mdcmd (319): spindown 9 Mar 15 20:17:16 Icarus kernel: mdcmd (320): spindown 1 Mar 15 20:17:19 Icarus kernel: mdcmd (321): spindown 3 Mar 15 20:35:55 Icarus kernel: mdcmd (322): spindown 5 Mar 15 20:50:02 Icarus kernel: mdcmd (323): spindown 11 Mar 15 20:50:05 Icarus kernel: mdcmd (324): spindown 6 Mar 15 20:52:32 Icarus kernel: mdcmd (325): spindown 10 Mar 15 22:42:33 Icarus kernel: mdcmd (326): spindown 4 Mar 15 22:46:15 Icarus kernel: mdcmd (327): spindown 2 Mar 15 23:00:01 Icarus kernel: mdcmd (328): check Mar 15 23:00:01 Icarus kernel: md: recovery thread: check P ... Mar 15 23:00:01 Icarus kernel: md: using 1536k window, over a total of 2930266532 blocks. Mar 16 12:34:01 Icarus shfs/user: err: shfs_rmdir: rmdir: /mnt/disk12/isos/plex/Library/Application Support/Plex Media Server/Cache/Transcode/Sessions/plex-transcode-f1e2b95f6bdc67ae-com-plexapp-android-e2adab5a-bb85-4588-adb2-e111aa18417c (39) Directory not empty Mar 16 13:12:53 Icarus kernel: mdcmd (329): spindown 9 Mar 16 17:04:44 Icarus kernel: mdcmd (330): spindown 9 Edited March 17, 2017 by pyrater added system specs Quote Link to comment
Frank1940 Posted March 17, 2017 Share Posted March 17, 2017 Post up your diagnostics file. Quote Link to comment
pyrater Posted March 17, 2017 Author Share Posted March 17, 2017 (edited) Ty for the reply, Attached! Last check completed on Fri 17 Mar 2017 11:30:13 AM MDT (today), finding 0 errors. Duration: 1 day, 12 hours, 30 minutes, 12 seconds. Average speed: 22.8 MB/sec icarus-diagnostics-20170317-1254.zip Edited March 17, 2017 by pyrater Quote Link to comment
JorgeB Posted March 17, 2017 Share Posted March 17, 2017 The hardware you're using will have some bus limitations, and your current config is not optimal, still that's way to slow. You can try this changes and see if it helps. The top 2 PCI-X slots and the bottom 2 share a PCI-X channel, so not optimal to use 2 controllers on the same channel, since you only have 14 HDDs remove one of the controllers and connect 7 disks each to a SAT2 on slot 1 or 2 and the other one on slot 3 or 4. Changes these tunables (Settings -> Disk Settings) md_num_stripes: 4096 md_sync_window: 2048 md_sync_thres: 2000 Run a new check and see if it improves. Quote Link to comment
Frank1940 Posted March 17, 2017 Share Posted March 17, 2017 It looks like Plex was doing something (I don't use Plex so I am not familiar with it's inter-workings) that spanned a wide time period while the server was apparently doing the parity check. If the Plex operation involved a lot of disk IO that could be part of your problem. Quote Link to comment
JorgeB Posted March 17, 2017 Share Posted March 17, 2017 25 minutes ago, Frank1940 said: It looks like Plex was doing something (I don't use Plex so I am not familiar with it's inter-workings) Yes, looks like it's scanning for media, also a lot of segfaults, this would explain the lower than expected speed even with a non optimal setup. Quote Link to comment
pyrater Posted March 18, 2017 Author Share Posted March 18, 2017 thanx guys you rock! Ill try the new settings and move the cards. Stopping plex isnt really an option so ill live with that. Quote Link to comment
JorgeB Posted March 18, 2017 Share Posted March 18, 2017 Stopping Plex is not really needed, just avoid scanning for new media during a parity check. 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.