May 25, 201214 yr Wasn't sure where to post but here goes. Running RC3 Had server running fine for quite some time (N36l 6 discs) decided to upgrade hardware. Built new server moved all disks over to it tonight and booted, all checked out ok with all disks accounted for a green lights and then tried to bring the array online. This took forever with the website not responding but being able to telnet. Eventually the web came back but then said 2 disks missing and a red ball (stupidly did not get syslog at this point was panicing). I rebooted the server and now have 1 disk as a redball (2). I have posted 2 syslogs I think there may be an issue with the controller AOC-SASLP-MV8. I tried bringing array online again as I know (thought parity was good) but still wouldn't come up and start rebuilding (see syslogs). I am starting to panic a bit as to what has gone wrong and what I may have lost. I don't really know if disk 2 actually has anything wrong with it as all discs pass smart. Motherboard has 6 ports so I am going to look at using them to try and eliminate the card as the syslogs are filled with May 26 05:19:22 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 05:19:22 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 05:19:22 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 05:19:22 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 05:19:22 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 05:19:22 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 05:19:22 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 While it is sitting there trying to come online so I think it may be the issue (faulty??) Any help would be appreciated as I do not want to have lost my data. syslog.zip
May 25, 201214 yr This is an issue with your hardware. Could be: 1. Bad cable from controller card to hard disk. 2. Bad hard disk drive cage (if you're using those). 3. Controller card not seated properly in motherboard. 4. Bad motherboard. 5. Bad power. 6. Bad hard drive(s) (but unlikely if they came from a previously running server). Not an -rc3 issue. Best way to isolate bad h/w is by simplifying things. For example, set up an array with disks connected only to motherboard ports - verify all is well. Now move disks to only controller ports - verify all is well, etc. Have to use deduction, Watson Of course this is riskier if you don't have any extra drives lying around
May 25, 201214 yr Author Thanlsnfor the reply, does the red ball still stand ie if I get the hardware working do I need to let the disk rebuild??
May 26, 201214 yr Author OK I think I have found the error and tested with a spare drive and usb I have. On the MV8 card it has 2 ports 1 marked 0-3 the other 4-7. When I have a drive connected to port 4-7 cable out 2 I get those errors in the syslog. May 26 06:12:57 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:12:57 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 06:13:00 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:13:00 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 06:13:04 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:13:04 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 06:13:06 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:13:06 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1100000 May 26 06:13:12 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:13:12 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 06:13:21 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:13:21 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 06:13:23 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:13:23 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 May 26 06:13:32 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2198:port 6 ctrl sts=0x199800. May 26 06:13:32 Tower kernel: drivers/scsi/mvsas/mv_sas.c 2200:Port 6 irq sts = 0x1000000 If I have drive or drives connected to any other port or cable everything is clean. I have swapped cables, drives, and norco drive cages but it appears as tho 1 port out of the 8 on the card is faulty and it stays with the port. Now I am going to plug the 6 drives into the 6 onboard sata slots. My red ball do I unassign the drive reboot reassign and let parity rebuild it or is there something else I should do.
May 26, 201214 yr Author OK so before I do this just to confirm this is the procedure 1 Turn on server array comes up with red disk and array stopped. 2 go to the disk with red globe and select "no device" 3 Restart server 4 when server comes back up for disk 5 (what will it say missing disk??) select the disk that was originaly failed 5 start array and let disk build from parity.
May 26, 201214 yr OK so before I do this just to confirm this is the procedure 1 Turn on server array comes up with red disk and array stopped. 2 go to the disk with red globe and select "no device" 3 Restart server 4 when server comes back up for disk 5 (what will it say missing disk??) select the disk that was originaly failed 5 start array and let disk build from parity. yes.
May 26, 201214 yr Author Thanks guys all back up. Had to bring the array online with the drive removed then offline it then reasign the disk but all good rebuilt syslog clean no errors. Next time test hardware before assuming it's good.
Archived
This topic is now archived and is closed to further replies.