June 8, 200719 yr I've been running my unraid server for about a week now while migrating drives and data to it from another system. I haven't noticed any problems except that hdb (disk1) keeps spinning up for no aparent reason. Nothing will be accessing any of the drives and it will go from flashing green to solid green in the web gui. I didn't think much of it, but decided to check the syslog anyhow. I don't understand most of what's recorded in the log, but I noticed a lot of errors and it made me concerned. I have attached the syslog and I'm hoping someone can tell me what it means.
June 8, 200719 yr Author A side note to the drive spinning up for no aparent reason. I just noticed the drive was now showing a flashing green indicator again and it has been far less than the hour I designated for drive spin down delay under settings.
June 8, 200719 yr The "can't shrink filesystem on-line" error in your log seems to be a harmless side effect of the remount command, issued to each drive after the 'recovery thread' is awakened. The 'recovery thread' seems to awake when a parity sync or similar operation may be needed, such as after a new drive is added. I'm guessing a normal file system mount compacts the drive structures while loading them, but a remount can't, because they are still in use. A Linux expert would have to verify that. I found this error a number of times in various syslog's. The other error sequence: hdf: drive_cmd: status=0x51 { DriveReady SeekComplete Error } hdf: drive_cmd: error=0x04 { DriveStatusError } ide: failed opcode was: 0xb0 smart ioctl: Input/output error occurred 8 times, randomly spaced, for the pair of Maxtor IDE drives hdf and hde, during a period of about 10 minutes immediately following the above 'recovery' operation, then stopped and never occurred again, in this log. The only other place I've seen the exact same sequence is in a log file in this (probably unrelated) thread http://lime-technology.com/forum/index.php?topic=741.0. I've seen a similar sequence referring to my only IDE drive, also a Maxtor. I have personally never seen them refer to SATA drives, only IDE. I realize this does not really help, but perhaps there's enough additional info for one of the experts. Everything else seemed normal in your log, nothing unusual referring to hdb, your disk1. Are you sure you don't have a program like Picasa or similar that does an occasional scan of media folders, perhaps one of which is located on your disk1?
June 8, 200719 yr The error messages are a result of the hard drives on /dev/hde and /dev/hdf not supporting SMART (or SMART not enabled on the drives). This could be a bios setting. The full syslog following a boot would be nice to see. What are you using for IDE disk controller?
June 10, 200719 yr Author I do have some drives that don't display SMART info. The funny thing is that they are the same make as others I have that do show SMART data. Also, the ones that give the info are connected to he same controller as ones that don't. I wonder why like drives would be different on the same controller.
June 10, 200719 yr My oem asus mb, has no smart enabling. Workaround is to do a telnet: cat /proc/ide/hda/smart_values cat /proc/ide/hdb/smart_values .... if it works, add to GO script /rene
June 10, 200719 yr RockDawg, I'm curious how you added that red outline with rounded corners in your screen image. Very nice way to highlight!
June 13, 200719 yr Author I used Paint just like dschur said. Tom - Once my server finishes copying a bunch of files, I'll post a full boot log.
June 17, 200719 yr Author The error messages are a result of the hard drives on /dev/hde and /dev/hdf not supporting SMART (or SMART not enabled on the drives). This could be a bios setting. The full syslog following a boot would be nice to see. What are you using for IDE disk controller? I have attached a log that was taken immediately after a boot. The missing SMART info seems odd because, as I said earlier, it's across different controllers and some of the drives that don't show the info are the exact same model as ones that do. I'm using two IDE controllers. A Promise Ultra 100TX2 and a Hightpoint Rocket 133.
June 20, 200719 yr did you try: cat /proc/ide/hda/smart_values It worked for me, but im not sure if you are chasing the solution+reason, or can live with solution+no reason. /rene
June 21, 200719 yr Author did you try: cat /proc/ide/hda/smart_values It worked for me, but im not sure if you are chasing the solution+reason, or can live with solution+no reason. /rene Sorry, I forgot to try that from your earlier post. Anyhow, I did try it now and this is what I get: 0010 2703 c900 91c9 0026 0000 0000 3204 fd00 2efd 0000 0000 0000 3305 fd00 00fd 0000 0000 0000 0106 fd00 00fd 0000 0000 0000 0a07 fd00 00fc 0000 0000 0000 2708 fa00 29e6 00e2 0000 0000 3209 cb00 e9cb 00f5 0000 0000 2b0a fd00 00fc 0000 0000 0000 2b0b fd00 00fc 0000 0000 0000 320c fd00 e8fd 0000 0000 0000 32c0 fd00 00fd 0000 0000 0000 32c1 fd00 00fd 0000 0000 0000 32c2 1d00 16fd 0000 0000 0000 0ac3 fd00 73fc 0038 0000 0000 08c4 fd00 00fd 0000 0000 0000 08c5 fd00 00fd 0000 0000 0000 08c6 fd00 00fd 0000 0000 0000 08c7 c700 00c7 0000 0000 0000 0ac8 fd00 00fc 0000 0000 0000 0ac9 fd00 00fc 0000 0000 0000 0aca fd00 00fc 0000 0000 0000 0bcb fd00 00fc 0000 0000 0000 0acc fd00 00fc 0000 0000 0000 0acd fd00 00fc 0000 0000 0000 2acf fd00 00fc 0000 0000 0000 2ad0 fd00 00fc 0000 0000 0000 24d1 f000 9ff0 0000 0000 0000 32d2 fd00 00fc 0000 0000 0000 32d3 fd00 00fc 0000 0000 0000 32d4 fd00 00fd 0000 0000 0000 0082 0656 5b01 0003 0001 5202 0000 0000 0000 0000 0000 0000 003b 0000 fdf0 020f 0011 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0035 ff8c 0073 0030 0000 8800 17bd 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0600 b300 I get this type of response from any drive I run the command on. No mattrer whether it reports SMART info in unraid or not.
June 21, 200719 yr Author OK, I see now. The response in the telnet window was meaningless. After runnign the command and then returning to the unraod web gui and refreshing I can now see the temps for drives I ran the command on that weren't showing the data before. That's cool! Now what is this GO script you mention?
June 21, 200719 yr in the /config directory on the flash is a file "go", open as a text file and add the "cat" commands for the drives necessary. voila, next time you boot you got the SMART info. Why it is needed I have no clue. :-) /Rene
June 23, 200719 yr Author IT looks like simply running the command once from telnet did the trick. I couldn't find the go file through the network, so I shut down the server and put the thumb drive in my desktop. I then found it, but realized I didn't write down which drives I needed to run the command for. So I put the thumb drive back in the server and rebooted and all the drives were still showing a temp. Subsequently, I have rebooted multiple times including even disconnecting the power cord for a few minutes and each time the web GUI shows the temps for all my drives. It would appear that simply running the command from telnet did the trick since I never altered my go file. Parsec, if it's not too much trouble, could you remove those lines from your go file and see what happens? I'm just curious if I just got lucky somehow or that is indeed all that needs to be done. I'd like to know so I can put that in the wiki.
June 28, 200719 yr Author Has anyone else experienced this? My temps still show and I have yet to enter anything into the go file? Why is it that simply running the command once would make a permanent change. What exactly does the command do?
June 28, 200719 yr Sorry, I were away on business, checking out a new jobopp, browsed houses and cars. Removed the cat-commands from go script, rebooted from within the webinterface, and now only get temp from SATA drives on Promisecontroller, but no temp from IDE drives on MB embedded controller. So the cat-effect did not survive a reboot.
June 29, 200719 yr Author Sorry, I were away on business, checking out a new jobopp, browsed houses and cars. No bother. That sounds like a goof thing. I hope it turns out well for you whatever you choose. Removed the cat-commands from go script, rebooted from within the webinterface, and now only get temp from SATA drives on Promisecontroller, but no temp from IDE drives on MB embedded controller. So the cat-effect did not survive a reboot. Thanks for the reply. I wonder why it "stuck" on mine?
September 15, 200718 yr Author My oem asus mb, has no smart enabling. Workaround is to do a telnet: cat /proc/ide/hda/smart_values cat /proc/ide/hdb/smart_values .... if it works, add to GO script /rene Is there a command like this for SATA drives?
September 15, 200718 yr Perhaps what's happening is that the drives didn't have SMART enabled. The 'cat' command is something like the DOS 'type' command. Accessing files in the '/proc' directory is special - you're not really accessing real files, but instead using a "pseudo" file sysystem built into linux to access system information (google 'linux proc file system' if you're curious). Probably what's happening is that the ide driver handling 'proc' access is enabling SMART before outputting the values. I'll look into adding automatic 'SMART' enable.
September 16, 200718 yr Author In the mean time, since that command worked for me before, is there any way to run it on a SATA drive?
September 29, 200718 yr Author In the mean time, since that command worked for me before, is there any way to run it on a SATA drive? Sorry to bump this, but I'd still really like to know if there's a similar command to run on SATA drives. Can anyone help?
Archived
This topic is now archived and is closed to further replies.