aRe-eYe-pEa Posted March 11, 2011 Share Posted March 11, 2011 This maybe me being an idiot which happens more often than not....... :'( I had a drive fail on me. So, I tried to rebuild it with a new hard drive. It failed to format so I powered down and then added a different disk. When I did this the array came back online and I had 2 hard drives that were "unformatted" besides the one that had failed. All my initial drives were started with v4.6 that didn't have the 4k alingment option. I then upgraded to v4.7 and selected 4k alignment. These 2 drives were Seagate 1.5TB's that didn't need the 4k. After taking them out of my server I found that they had been set to sector 64 or 4k alignment from info from Get Data Back. So, I could have used Joe L's utility to fix it. I was just wondering if anyone else has had this issue? Is this something that I should have been smart enough to aviod? Quote Link to comment
draz Posted March 21, 2011 Share Posted March 21, 2011 Hi, I recently upgraded from 4.5.x to 4.7. Following the instructions in the readme in the release and copying the three files to the flash drive of my already running NAS. After a reboot, I have two disks with a red dot and two with a blue dot. The array was running fine pre upgrade but now won't start due too too many missing or disabled disks. I have seven drives and all seven are showing in the devices tab. I've tried unassigning and re-assigning all the disks to no avail. All drives including red and blue are showing their SMART data, temps and sizes. Is it possible to provide some more useful log files that may help indicate what is occuring. I am not sure what log files would be most useful. Regards, Draz. Quote Link to comment
Joe L. Posted March 21, 2011 Share Posted March 21, 2011 Hi, I recently upgraded from 4.5.x to 4.7. Following the instructions in the readme in the release and copying the three files to the flash drive of my already running NAS. After a reboot, I have two disks with a red dot and two with a blue dot. The array was running fine pre upgrade but now won't start due too too many missing or disabled disks. I have seven drives and all seven are showing in the devices tab. I've tried unassigning and re-assigning all the disks to no avail. All drives including red and blue are showing their SMART data, temps and sizes. Is it possible to provide some more useful log files that may help indicate what is occuring. I am not sure what log files would be most useful. Regards, Draz. post a system log. Attach it to your next post. ZIP it if it is too big to attach otherwise. Instructions under "troubleshooting" in the wiki. Quote Link to comment
redplanet Posted March 22, 2011 Share Posted March 22, 2011 Just to mention - I'm another user who hit the HPA problem when upgrading from 4.6 to 4.7. On both my unraids, "hdparm -N pnnnnnnn /dev/..." fails with "input/output error" (including on first try after power-up). I've seen all the clever stuff posted here about how to fix it, but I just haven't the time and I really don't fancy the risk of losing data either. Nor replacing my two GigaByte MBs. So I'm just here to vote for a 4.7.1 release which might fix the problem without me having to take my chances and start tinkering. I know, I know, you need to get 5.0 out of the way first. But after that... please? Pretty please? Quote Link to comment
PeterB Posted March 23, 2011 Share Posted March 23, 2011 ... I really don't fancy the risk of losing data either. Nor replacing my two GigaByte MBs. Have you managed to disable the HPA function in the mobo bios? If not, you continually run the risk of losing data! So I'm just here to vote for a 4.7.1 release which might fix the problem No changes in unRAID can 'fix' the problem. The only true fix is to disable (or remove) the HPA functionality in the bios, and remove the drive corruption introduced by that. unRAID can be altered to ignore the problem, but it won't, necessarily, prevent data corruption from being caused by HPA. Quote Link to comment
artagnan Posted March 24, 2011 Share Posted March 24, 2011 Hi! I follow exactly all the steps to make my 4GB Scandisk Cruzer USB a bootable one (with Win7), but when I right-click on the file 'make_bootable' and select 'Run as administrator' a DOS windows is opened, gives me the Volume name (it is UNRAID) and the Volume Serial Number and finally gives me that Error: Error: volume label must be UNRAID Press any key to continue... What does I make wrong? Thanks in advance. Quote Link to comment
Rajahal Posted March 24, 2011 Share Posted March 24, 2011 Are you sure the flash drive is labeled 'UNRAID' in all capital letters? unRAID, or unraid, or anything else won't work. Quote Link to comment
artagnan Posted March 24, 2011 Share Posted March 24, 2011 Yes, UNRAID in capital... I have test with 2 different USBs. Quote Link to comment
S80_UK Posted March 25, 2011 Share Posted March 25, 2011 Just in case - try changing the volume label in Windows (right click on drive, Properties) and make sure that there are no leading or trailing spaces in the label. Quote Link to comment
artagnan Posted March 25, 2011 Share Posted March 25, 2011 Thanks S80_UK, but I tried that before. Make a bootable Unraid USB seemed to be easy, but there's no way to do it with the "make_bootable.bat", always the same Error... And I've downloaded the 4.7 from the Unraid web. I've downloaded the 4.6 and test with its "make_bootable.bat". Is there any other way to make the USB bootable? Quote Link to comment
SSD Posted March 25, 2011 Share Posted March 25, 2011 These announcement threads are for people that have specific problems with new releases. If you are in need of support, please create a new thread in the General Support forum. I think you may find more people willing to help there. Quote Link to comment
artagnan Posted March 25, 2011 Share Posted March 25, 2011 I'll do it, but I'm having problems installing the 4.7 firm, so I thought here was the right place to ask. Quote Link to comment
spasszeit Posted April 15, 2011 Share Posted April 15, 2011 I seem to have been having problems with 4.7 which I did not have before. I run two servers and ever since upgrading to 4.7 I get occasional drive drive drop off from the array. First time it happened I thought the drive was bad. Re-seated the cables, that did not help - during the system post I see one port not being detected. When array starts the disk is missing. I replaced the drive with a new one and the array rebuilt itself just fine. I put the bad drive into the other tower and ran pre-clear on it - no smart errors. Last night I had the same problem on the other server. The problem exhibited itself by inability to access web interface, though access to shares via windows and via Telnet worked fine. I captured the syslog - see attached. There is an awful lot of messages related to, I think, my SM AOC-SASLP-MV8 and a disk. I run two AOC-SASLP-MV8 cards in each server. Now, my friend built a server recently as well. He is running one AOC-SASLP-MV8 card, and he mentioned to me that ever since upgrading to 4.7 he has also experienced a problem like mine. I hope Tom can look into this and advise on what's going on here and how we can address it. Syslog_Apr_14.zip Quote Link to comment
bcbgboy13 Posted April 15, 2011 Share Posted April 15, 2011 Before we derail this to 4.7 and the SASLP I think we need more info that was not provided in the partial syslog. When AOC-SASLP-MV8 started shipping it came originally with firmware .15N The new ones are with firmware .21N and this firmware if bigger in size than the original one. A user here has provided picture from the new firmware menu at boot-up time and it is clearly visible that SM added a RAID capabilities to the "new" card. Perhaps the RAID overhead now build into the card can cause this loop in initialization of a "sleeping or idle" drive. The firmware revision is visible at boot-up time and I believe is identifiable in the syslog: firmware .15N - identified as "0x101" firmware .21N - identified as "0x103" Syslog examples: Dec 14 13:57:08 Tower kernel: mvsas 0000:02:00.0: mvsas: driver version 0.8.4 Dec 14 13:57:08 Tower kernel: mvsas 0000:02:00.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19 Dec 14 13:57:08 Tower kernel: mvsas 0000:02:00.0: setting latency timer to 64 Dec 14 13:57:08 Tower kernel: /usr/src/sas/trunk/mvsas_tgt/mv_spi.c 491:Init flash rom ok,flash type is 0x101. I can be wrong but this is worth investigating IMHO. And perhaps a simple downgrade to the older firmware can cure this problem. Quote Link to comment
spasszeit Posted April 15, 2011 Share Posted April 15, 2011 I downloaded the full log, there are two entries: Apr 14 23:05:01 Tower2 kernel: mvsas 0000:03:00.0: mvsas: driver version 0.8.4 Apr 14 23:05:01 Tower2 kernel: mvsas 0000:03:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 Apr 14 23:05:01 Tower2 kernel: mvsas 0000:03:00.0: setting latency timer to 64 Apr 14 23:05:01 Tower2 kernel: /usr/src/sas/trunk/mvsas_tgt/mv_spi.c 491:Init flash rom ok,flash type is 0x101. Apr 14 23:05:01 Tower2 kernel: mvsas 0000:02:00.0: mvsas: driver version 0.8.4 Apr 14 23:05:01 Tower2 kernel: mvsas 0000:02:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 Apr 14 23:05:01 Tower2 kernel: mvsas 0000:02:00.0: setting latency timer to 64 Apr 14 23:05:01 Tower2 kernel: /usr/src/sas/trunk/mvsas_tgt/mv_spi.c 491:Init flash rom ok,flash type is 0x101. Both have the same (old) firmware. Quote Link to comment
bcbgboy13 Posted April 15, 2011 Share Posted April 15, 2011 Well in this case I agree that something has to be done. Quote Link to comment
Rajahal Posted April 15, 2011 Share Posted April 15, 2011 Are you sure your power supply is powerful enough to handle all the drives you are running? If you have two servers I'm guessing that at least one of them is full. Quote Link to comment
spasszeit Posted April 15, 2011 Share Posted April 15, 2011 Are you sure your power supply is powerful enough to handle all the drives you are running? If you have two servers I'm guessing that at least one of them is full. The tower where one drive dropped off last night has 13 drives, most are green. Running on a 650W high efficiency PSU, well below its capacity. Quote Link to comment
Rajahal Posted April 15, 2011 Share Posted April 15, 2011 But is it a single rail PSU? That matters far more than the wattage. There are 850 W PSUs with multiple rails that can't handle as many drives as a 500 W PSU with a single rail... If you aren't sure, just post the make and model of the PSU and we can look it up. Quote Link to comment
spasszeit Posted April 16, 2011 Share Posted April 16, 2011 But is it a single rail PSU? That matters far more than the wattage. There are 850 W PSUs with multiple rails that can't handle as many drives as a 500 W PSU with a single rail... If you aren't sure, just post the make and model of the PSU and we can look it up. Single rail. Quote Link to comment
Carabolic Posted April 26, 2011 Share Posted April 26, 2011 Two Questions for Joe.L: My unRaid server consist of several 1.5TB drives. As I need to expand the storage capacity, I'm planning to buy and insert a 2TB drive as the Parity drive and shifting that old 1.5 TB drive as a new storage drive. This will give me a 2TB 4k drive on the parity and 1.5 TB 512k drives for the actual storage. (my plan is to shift down the old 1.5 TB drives further as I will replace the top ones by time with 2TB ones. 1. When inserting the 2TB drive as parity, do I need to run the pre-clear script before creating the new parity data? 2. I think I read a couple of years ago that I could insert an empty parity drive and attach the old parity drive in the system and have the old parity data to be transfered to the new disk. Is this correct? Regards, Peter Quote Link to comment
Joe L. Posted April 26, 2011 Share Posted April 26, 2011 Two Questions for Joe.L: My unRaid server consist of several 1.5TB drives. As I need to expand the storage capacity, I'm planning to buy and insert a 2TB drive as the Parity drive and shifting that old 1.5 TB drive as a new storage drive. This will give me a 2TB 4k drive on the parity and 1.5 TB 512k drives for the actual storage. (my plan is to shift down the old 1.5 TB drives further as I will replace the top ones by time with 2TB ones. 1. When inserting the 2TB drive as parity, do I need to run the pre-clear script before creating the new parity data? No, you do not "need" to run the pre-clear script on the new drive, but since 1 out of 5 drives seems to fail the preclear tests lately, all I can ask is "do you feel lucky?" If you would rather not trust your luck, I would suggest that before you assign the new drive to your array that you put it through at least one preclear cycle. 2. I think I read a couple of years ago that I could insert an empty parity drive and attach the old parity drive in the system and have the old parity data to be transfered to the new disk. Is this correct?You are mixing two different concepts. To replace a parity drive you just remove the old drive, plug the new drive onto the same cable, and start the array. Parity will be calculated and stored on the parity disk. once it is complete you should then perform a parity check to be certain it can be read. Regards, Peter Quote Link to comment
Carabolic Posted April 28, 2011 Share Posted April 28, 2011 Two Questions for Joe.L: My unRaid server consist of several 1.5TB drives. As I need to expand the storage capacity, I'm planning to buy and insert a 2TB drive as the Parity drive and shifting that old 1.5 TB drive as a new storage drive. This will give me a 2TB 4k drive on the parity and 1.5 TB 512k drives for the actual storage. (my plan is to shift down the old 1.5 TB drives further as I will replace the top ones by time with 2TB ones. 1. When inserting the 2TB drive as parity, do I need to run the pre-clear script before creating the new parity data? No, you do not "need" to run the pre-clear script on the new drive, but since 1 out of 5 drives seems to fail the preclear tests lately, all I can ask is "do you feel lucky?" If you would rather not trust your luck, I would suggest that before you assign the new drive to your array that you put it through at least one preclear cycle. 2. I think I read a couple of years ago that I could insert an empty parity drive and attach the old parity drive in the system and have the old parity data to be transfered to the new disk. Is this correct?You are mixing two different concepts. To replace a parity drive you just remove the old drive, plug the new drive onto the same cable, and start the array. Parity will be calculated and stored on the parity disk. once it is complete you should then perform a parity check to be certain it can be read. Regards, Peter Ok, Thanks Joe! Every thing's running fine now! One last question... As I've changed the filesystem setting to 4K sectors, will there be any problems when I attach a new 512k drive (my former parity drive) after I've removed any partitions on it and attached it as a file drive into the array? Will unRaid try to use it as a 4K or a 512K drive after I've changed the default setting to 4K? //Peter Quote Link to comment
lionelhutz Posted April 28, 2011 Share Posted April 28, 2011 As I've changed the filesystem setting to 4K sectors, will there be any problems when I attach a new 512k drive (my former parity drive) after I've removed any partitions on it and attached it as a file drive into the array? Will unRaid try to use it as a 4K or a 512K drive after I've changed the default setting to 4K? There is no such think as a 512b or 4k drive (not yet anyways). The setting just changes the location of the disk partition. Every drive presently available will work fine using the 4k-aligned setting except a WD EARS drive with the alignment jumper. 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.