February 27, 200917 yr I just experienced a major problem yesterday. 1 disk is red and saying now installed. 1 disk is gray and saying not installed. What is the best approach to take as I do not want to lose this data? Thanks for any help!!!
February 27, 200917 yr I just experienced a major problem yesterday. 1 disk is red and saying now installed. 1 disk is gray and saying not installed. What is the best approach to take as I do not want to lose this data? Thanks for any help!!! First step, before you do anything, and BEFORE you reboot, post a copy of your syslog. Then, proceed only after advice based on the syslog contents. See this link for how to capture the syslog Do NOT press the "restore" button. It does not restore data. See here: Joe L.
February 27, 200917 yr Author Hi Joe- Unfortunately, I already shutdown and rebooted. Would the syslog still help? I was able to copy the data from disk 1 to an external, but not disk 13. Thanks, Jarrad
February 27, 200917 yr Hi Joe- Unfortunately, I already shutdown and rebooted. Would the syslog still help? Thanks, Jarrad It will still help, since the drives are still not being seen, but we might not (read probably won't) be able to tell you WHY they failed and what may have caused it.
February 27, 200917 yr Unless there was a serious electrical spike, it is extremely rare for more than 1 drive to fail at the same time. Usually it means something in the machine had a problem, such as a common channel or controller shared by the 2 drives, and there is nothing at all wrong with the drives. A reboot often 'fixes' the problem, but it really really helps to have that syslog, *before* you reboot. Another common case is when 2 drives are on an IDE channel, real or emulated, and one drive has serious problems. If the kernel has to send a reset to the drive in an effort to recover it, the reset actually goes to the channel, causing both drives on the channel to try to reset, and I've often seen both drives then have trouble responding to the reset, when really only one of the drives had serious issues. (This is one more reason why I prefer AHCI mode to IDE Emulated mode for SATA drives, since the emulation actually seems to emulate one of the flaws of the IDE channel scheme!) Now that you have rebooted, are there still 2 drives down? Some times, unRAID just needs a 'reset' to tell it problem is over and array is fine, see the Trust My Array procedure. But as he said, we still recommend a check of a current syslog first, to make sure.
March 3, 200917 yr Author I took the weekend off and turned the machine back on this morning hoping for some miracle. The first disk and the 13th disk are both blue balls. So what's my best approach now? I was able to off load all of the data from disk 1 onto an external last week, but I really need to get the data back from disk 13. Is there an easy way to do this? (I tried pulling the drive and connecting to a Vista machine, but it said that the format was not recognized...or something like that) I think the error might have been caused by me pulling out a disk without disconnecting it first from the devices tab. (that's the only thing that I can think of) Thanks again for any and all help!!!
March 3, 200917 yr I've attached a syslog to help The syslog indicates you have apparently replaced disk1, and that disk13 is a new disk (at least to unRAID it is new since you last rebooted.) Your forum etiquette inexperince is showing. Please do NOT open up many threads in different forums for a single problem. Most of us will not follow you around to try to learn all the pieces of the puzzle to help you. I will discontinue any responses in your other thread now. Perform the steps indicated and report back. It will provide the clues needed to direct you further. It might be easy to recover. Did you discover a reason why you had two disks go offline at the same time yet? Was it cabling? Or still a mystery? Joe L.
March 3, 200917 yr Because the array never started, I cannot tell if there are any mounting issues with Disk 13, but otherwise there are no issues at all with it, so I assume it will be fine once the array can be started. As to why it shows as a new drive now, I can't tell without checking previous syslogs. Something must have gone wrong in the past, and it was dropped from the array. On the other hand, Disk 1 is REALLY messed up, claiming to be a Seagate ST_M13FQBL drive, SATA II, of 4GB! That is 4 gigabytes only! Plus it is reporting numerous device and other errors right from the start, looks probably unreadable. Since you say you were able to copy the data off this, it may have just had a meltdown? Or this was just a very bad boot? Looks unusable at the moment, needs to be pulled. Try a total power off, then boot up, and if the syslog does not look better for Disk 13, sdm at ata12.00, then it has to be pulled. I did a Google search for ST_M13FQBL, and found a few reports of good Seagate drives suddenly reporting their model name as that, possibly associated with a bad firmware update, with no resolution except to return the drive. Looks to me like corrupted firmware. After you pull Disk 1, you will have to click the Restore button before you can start the array, and it will need to rebuild parity from the remaining drives. At that time, Disk 13 should be readable. Very minor issue department: your 2 Seagate terabyte drives probably still have their jumpers installed, see the Improving unRAID Performance wiki page, Remove SATA150 Jumper section.
March 3, 200917 yr Author Hi Joe- I'm sorry for opening a new thread! I just did the vol_id /dev/sde1 as described below and here is the output (this was for disk 13...disk 1 returned nothing as it's a new drive): ID_FS_USAGE=filesystem ID_FS_TYPE=reiserfs ID_FS_VERSION=3.6 ID_FS_UUID=942baf70-1e5a-4d1e-8503-ef3530561dab ID_FS_UUID_ENC=942baf70-1e5a-4d1e-8503-ef3530561dab ID_FS_LABEL= ID_FS_LABEL_ENC= I then did the reiserfsck -q -y /dev/sdf1 and it returned "no corruptions found." I unassigned disk 1, removed the drive, and it has a red ball next to it on the main section. Disk 13 is the disk with all the data and it has a blue ball. I tried to start it, but it stopped with "invalid expression" and now the "start" button is grayed out. At this point, it would be great if I could just start the system up and have Disk 13 and its data up and running, but if I have to I can move all of that data onto an external and then treat disk 13 like a new disk. (though more time consuming) What should I do next? Thanks again Joe! ------------------------------------------------------------------------------------------------------------------ If a disk is "blue" then it indicates it is not currently known in the superblock defining the array. It "might" have a file-system on it, it might not. It might have files on it, it might not. If the disk is functional, and has a file-system, then it is very easy to get the data off of it. Believe it or not, if the disk has a valid file system, the easiest way to proceed is to start your array. But first, let's make sure the file-system exists. First things, first... Un-assign the "blue" drives from the array. Before you do, make note of the linux "device" names shown in parentheses on the devices page. It will be (hd?) or (sd?) depending on if it is an IDE drive, or an SATA drive. LEt's assime for the rest of my examples, it is "sdf" Log in via telnet. Then type: vol_id /dev/sdf1 Notice I used sdf1, with a trailing "1", to reference the first partition on the disk. If you have a reiserfs file-system, it should return something like this: vol_id /dev/sdf1 ID_FS_USAGE=filesystem ID_FS_TYPE=reiserfs ID_FS_VERSION=3.6 ID_FS_UUID=ebc0aeb3-a2e7-4ee7-b4f2-cdf4e556abf6 ID_FS_UUID_ENC=ebc0aeb3-a2e7-4ee7-b4f2-cdf4e556abf6 ID_FS_LABEL= ID_FS_LABEL_ENC= Assuming it has a reiserfs file-system, you are in good shape. If it reports no file system, then post back here with what you do see. You can also post the output of fdisk -l /dev/sdf if it reports no file-system exists. (It shows how the disk was partitioned) If there is a file-system, then we can check it to see if it has any corruption: Type: reiserfsck -q -y /dev/sdf1 With any luck, the file-system will not have any corruption. If it does, the output should direct you how to proceed. Post the output here for guidance. Assuming a good file-system, just re-assign the drives on the management console and start the array and you can get to your files. (You may need to check the I'm sure box below the "Start" button) Joe L.
March 3, 200917 yr Author Hi RobJ- I pulled Disk 1 from the server. It has not installed and a red ball. Disk 13 has a blue ball, but has all the data on it that I need to keep. Are you saying that I should hit "restore" and it will bring disk 13 online with all the data? (I'm always leery of pressing the restore button after reading numerous threads) Thanks again!
March 4, 200917 yr Once you pull a drive, you don't have a choice, because parity is invalid. Yes, clicking the Restore button at the wrong time loses your parity info, but in your case, it's already unusable once the other drive is gone. Now there *is* one other choice here. Assuming you have a replacement drive for Disk 1 AND assuming Disk 13 is fine (that appears to be possibly questionable from your other posts?), you could try a new procedure that is a hybrid of the drive rebuild and the Trust My Array procedure. You would use the following invalidslot command: mdcmd set invalidslot 1 instead of the 99 version in that procedure. And in the last step, a drive rebuild of Disk 1 would begin, instead of a parity check. Disk 13 has a blue ball, but has all the data on it that I need to keep. Do you mean you have been able to mount it and verify that all your data on it is fine? UnMENU would be very helpful here, to let you view and check it.
March 4, 200917 yr Author Hi Robj- I was able to copy all the data from disk 1 to an external. Drive 13 seems to be fine, but I need to ensure the data stays on that drive instead of erasing it. I also have a new replacement drive for disk 1 that has yet to be formatted. So would that hybrid technique be my best bet? Is there a guide for steps to follow? Thanks
March 5, 200917 yr Author Can anyone else help? My Disk 13 contains some critical data, but it will only show up as a blue ball. How can I extract the data from that drive or how can I start unRAID without over writing the data. Thank you very much!
March 5, 200917 yr I just read this thread and am totally confused. I will try to give some facts. Consider them and then post back with any questions. If 1 drive fails, you can replace it and unRAID will rebuild it. If 2 drives fail, you lose the data on both drives. If unRAID crashes it will sometimes lose the file that keeps track of what drives are in the array (a file called super.dat). There is a procedure called "Trust parity" (look on the FAQ for the link). This will cause the array to put the array back together again and trust that parity is correct. It is important that the array really be "trustworthy" if you are going to do this. If unRAID crashes, fogets the array definition, and ONE of the disks fails at the same time, you can run a modified "trust" procedure, where instead of telling unRAID that the array is 100%, you can tell it that one data disk needs to be rebuilt. Hope this helps. Post back with any questions.
March 5, 200917 yr Author Yeah sorry it's so confusing! I turned on unRAID last week and one disk was gray and one disk was red. I was able to move the data off of drive 1 onto an external drive. (I have since removed that drive and added a new drive in it's place, but can return the old drive if necessary) Drive 13 on the other hand now shows up in the system as being blue. I need to figure out a way to either have the system come back online and have the data from drive 13 be safe or I need to move the data from drive 13 to an external. Does that make more sense? Basically, how can I get the data from drive 13 to come online or how can I move it to an external drive since it shows up as a blue ball? Thanks
March 5, 200917 yr Yeah sorry it's so confusing! I turned on unRAID last week and one disk was gray and one disk was red. I was able to move the data off of drive 1 onto an external drive. (I have since removed that drive and added a new drive in it's place, but can return the old drive if necessary) Drive 13 on the other hand now shows up in the system as being blue. I need to figure out a way to either have the system come back online and have the data from drive 13 be safe or I need to move the data from drive 13 to an external. Does that make more sense? Basically, how can I get the data from drive 13 to come online or how can I move it to an external drive since it shows up as a blue ball? Thanks You have checked disk13 and it has a valid reiserfs file-system. To get back online all you should need to do is: -- First... Since parity is already invalid, you might want to un-assign the parity drive on the devices page BEFORE you press the "Restore" button. The parity drive will just slow things down while you copy your critical data off the disk13. You can stop the array, re-assign the parity drive and re-start the array later once you have your data. -- Second, Press the "Restore" button to have unRAID store a new condifuration representing your existing drives. When you press the "Restore" button, all your data drives should turn blue. The "new" disk1 will probably show as unformatted. Very important. Disk13 should Not show as unformatted. If it does show unformatted, do not go any further. -- Third, You will then be given the opportunity to "Start" the array. You will need to check the "checkbox" under the start button first to enable it. When you start the array it should let you get to your disk shares and user-shares (if you have them enabled) Get your data copied where you think it might be safe. Now... you still have never answered my previous question. Did you ever identify the cause of your disks going off-line? Was it a loose cable as we originally suspected? Joe L.
March 5, 200917 yr Just a quick couple of additional facts ... 1. If a drive is red, unRAID has thrown it out of the array. unRAID will never use it again, although in certain cirumstances it will simulate it. 2. If a drive is blue, unRAID believes it is a NEW drive - not in the array but will be ADDED to the array if the array is starte. The only other time a drive is blue is if unRAID has detected a new disk that it will rebuild onto. The status message will tell you if that is the case. If you have already rescued the data from disk 1 (which you say you have), the other question is - is the data on disk 13 valid. If it is, parity doesn't matter - you can just redefine the array and get it to rebuild parity. But if disk 13 has failed, then you need your original disk 1 and your current parity to try to recover it. That is what Joe L. is trying to confirm I think. Too many chefs spoil the soup. Joe L. is the best at figuring out these things. Answer all his questions and he'll get you straight. Good luck!
March 5, 200917 yr He already verified disk13 as follows: I just did the vol_id /dev/sde1 as described below and here is the output (this was for disk 13...disk 1 returned nothing as it's a new drive): ID_FS_USAGE=filesystem ID_FS_TYPE=reiserfs ID_FS_VERSION=3.6 ID_FS_UUID=942baf70-1e5a-4d1e-8503-ef3530561dab ID_FS_UUID_ENC=942baf70-1e5a-4d1e-8503-ef3530561dab ID_FS_LABEL= ID_FS_LABEL_ENC= I then did the reiserfsck -q -y /dev/sdf1 and it returned "no corruptions found." I unassigned disk 1, removed the drive, and it has a red ball next to it on the main section. Disk 13 is the disk with all the data and it has a blue ball. I tried to start it, but it stopped with "invalid expression" and now the "start" button is grayed out. At this point, it would be great if I could just start the system up and have Disk 13 and its data up and running, but if I have to I can move all of that data onto an external and then treat disk 13 like a new disk. (though more time consuming) What should I do next? Thanks again Joe! However... I just noticed he said /dev/sde1 on the vol_id and /dev/sdf1 on the reiserfsck command. I think disk13 is /dev/sdf1, but I might be mistaken. Stiltz, Please re-verify you tested the correct drive.!!!!!!!!!!!!!!!!!!!!!!!!! Joe L.
March 5, 200917 yr Author Hi Joe- Disk 13 is sde1. I mis-typed "sdf1" in the 2nd part where I said it had no corruptions. In my notes, I have disk 1 as "sdl" Here's what I plan to do. Please let me know if this is correct. 1) Restart unRAID (without Disk 13 and with a new disk 1) 2) Start the array and let it make disk 1 available 3) Then go to devices page and unassign the parity drive 4) Shut down unRAID 5) Put original disk 13 (with data) back in 6) Start unRAID 7) Move data from disk 13 to an external (or to the new empty disk 1) go to devices page and assign parity drive Will that work? Ha, sorry for all the noob questions, but I'm just trying to get out of this mess with all the data intact!
March 6, 200917 yr Joe- did I get the steps correct? Thanks No, I do not think you do. I outlined the steps earlier. You will not be able to start the array unless you use the "Restore" button. You did not mention it in your liist of steps.
March 7, 200917 yr Author I was able to get a new Disk1 back online and moved the old data back onto it. Now Disk13 is showing up as unformatted. What is the easiest way to get the data off that drive? Thanks
March 7, 200917 yr I was able to get a new Disk1 back online and moved the old data back onto it. Now Disk13 is showing up as unformatted. What is the easiest way to get the data off that drive? Thanks Do NOT press the "Format" button. I have no idea what you have done to get to that point. I thought you had verified that disk13 had a reiserfs on it? Joe L.
March 7, 200917 yr Shut down (stop the array) and power off, then restart. Once it boots, grab the syslog again and post it here. I'm having a little difficulty keeping every one's issues separate, so perhaps it is time to get a fresh view of what the syslog says, before trying to advise further.
Archived
This topic is now archived and is closed to further replies.