Dear All,
As a new user of a LS PRO DUO 1TB I've chosen Raid1 to avoid any data loss in case of one HDD will fail. I've looked in the manual for several times but didn't found anything about the failure case.
What will happen if one day one HDD will crash. Using Raid 1 the LS PRO should work with one drive and all data will be there, but what are the next steps ? I've tried to search the forum but didn't found a step by step guideline to bring my system back to normal life.
Can you please provide me with solutions for the following scenarios:
1. One HDD will fail during the warranty phase (Buffalo has to replace the HDD or will they only change the full device including loss of data ???)
2. One HDD will fail after the warranty phase (I've to buy a new HDD)
I'm living in Germany if this makes any difference for the warranty-case. Hope you understood my bad english.
Best regards,
Peter
Due to new CE redulations (well in Europe anyways) if one of the HDD's fails then the entire device will have to be replaced (this is down to the fact that there are "exposed" cables when you take the drives out) You can technically replace the drive itself and restructure the RAID array.
On other thing, since it is a software based raid it is always best to keep a backup of the data on a different medium. If the software raid fails then you lose all your data.
Then whats the point of the RAID setup and the idea that its mirrored, if the software fails then you cant recover your data?
Our 1TB is setup as RAID 1 ,Im assuming that if Disk 1 fails our data is safe on Disk 2 but its not looking that way.
So I bought this to help preserve my data, but the suggestion is to ALSO back it up on ANOTHER media source. Again Why did I buy this drive then?
Not trying to come of gruff, just confused.
Please accept my apologies for the late reply.
RAID has never been a substitute for a good backup regimen. RAID only allows some redundancy in case of a disk failure. If you're using the Linkstation as your primary storage, we recommend backing it up, i.e. to an external hard drive.
You still haven't answered the question. How does one recover when one drive fails?
Hello,
I would like to join the rant on this topic.
I just purchased a new 2TB Linkstation Pro Duo last week.
Funny this topic got started, because I was looking in the manual to figure out
how to replace one of the drives if one of the two fails.
(That is what Raid 1 is right?)
The whole reason I purchase this device was to be able to swap the drives.
Now it looks like that is not possible????????
I am sure not going to purchase another device to back this one up.
That was the whole reason for purchasing this one.
I would like to know from Buffalo or someone else, how do I change my 1TB drive it one fails.
If I can't get this answer, I am going to dump this unit on ebay before I load it down with
to much data.
My intentions are to not back this unit up, but replace the drives when a failure occurs.
Thanks,
Keith
A lot of people waitng for an answer on this one I guess. I also agree about now being told to purchase ANOTHER backup unit?
Could someone please explain to me also why I bought this one in the first place then?
I have the same concern.
I bought my LS Pro Duo to use as backup storage. If it's unreliable, what good is it? I need to count on it being there in case data on my working computers is lost for any reason. Now I'm supposed to get a backup for the backup???
LOL .... I got a ls-mini just to back up my pro-duo due to this problem. Fat chance of that happening. That doesnt work properly either and I still have not gotten a reply on that one other than... hmm I'll run a test.... over a month ago.
I'll unload these two and get ONE reliable box. maybe HP.
I have a 1TB Drive Station set up as RAID1. Could you please elaborate on your comment that RAID only allows "some" redundancy in case of a disk failure. I too was led to believe that should one of the drives in the station fail, the other drive would have a mirrored copy of all data backed up to the point where the failed drive stopped operating.
How is a failed drive replaced in the Drive Station? When it is replaced, how do you get the data from the non-damaged drive to the new drive? Will the drive station automatically go into a comparison routine and copy the data onto the new drive from the functioning drive that was not replaced?
Thanks,
Rowland
Edit: Some people have already answered specifically about the LS Pro Duo. I don't use a NAS, sorry. However Raid 01 software or hardware solution serve the same purpose. The following apply to both.
Raid 01 customer review of Buffalo product HERE. It is important to make a difference between a HDD in Raid failure, and a Raid controller failure.
When the physical HDD attached to a Raid controller fails:
1. switch off
2. remove and replace the dead HDD (both HD size must be identical on Raid 01)
3. close and switch on
4. the reconstruction (copy of the drive to the new one) is automatic
When a Raid controller fails and damage the MFT or disk header, then all the data is usually gone: disk 01 + disk 02
Recovery companies can retrieve the data on the platters using clean rooms, but that's not cheap. Some software can help in the case of a Raid failure. Some reading HERE
admsupport,
Thanks for your response.
I assume, since you state in your reply
"I don't use a NAS, sorry"
That your response is not an official one from "Buffalo" the company,
and for this particular unit.
I think we all understand the theory on how it should work,
but does it?
What I am looking for, since I spent money on this product
an "Official" response from "Buffalo" that should have been
explained in the user manual.
"How do you replace a bad disk drive"
Put yourself in my shoes.
I purchased this unit 2 weeks ago, and after joining this forum I find out
that I am supposed to back this unit up too!
This was stated in an earlier post by a "Buffalo Representative"...
One even stated for some reason that makes no sense to me
that the unit has to be returned for repair for drive replacement.
So I need to go buy another device, to backup the device I just purchased
to back that up too... Come on ! What a crock !
If you guys visit Buy.com, you will see WD has a USB 2.0 device
that is mirrored 1TB Drives for 1/2 the price of this unit.
They even have a video of a WD representaive showing
how easy it is to pop open the cover and replace the bad disk drive.
I bet there manual is complete...
I have a number of computers on my network, and I purchased this unit
to be the destination of my data so I don't have to backup.
Micro Center recommended this unit based on my requirements.
They sold me one unit, not this one, and another one to back this one up...
I spent money on the LS Pro DUO that has the functionality of "Mirroring",
so if a failure occurs, I should be able to replace one of those drives
and do exactly as you explained myself.
Since I am a owner of the product, and you say you don't own one
I think you are "Assuming" what you say to be true.
I sure hope your right... but show me the documentation!
I would like to express my dissatisfaction of the product
and would like to demand that "Buffalo" provide instructions to
those of us who spent money on this unit.
I would like a PDF, that should have come with the purchase of my unit
telling me step by step instructions on how to replace a drive.
I wait anxiously wait for a response...
If there is no response from "Buffalo"...
I will also be posting reviews on this product on many websites sharing
the information that I became aware of after my purchase, warning
individuals to be careful in there selection of this product.
thanks for your intelligent response!
I was looking through the http://recovery-review.com/index.php?action=search link but they did not provide any tools that handle the xfs system recovery if the "software goes bad". The Pro Duo by default comes with an XFS filesystem and it looks like i will need some form of recovery, do you know of any tools that will unformat an XFS filesystem?
Thanks for the tip, i was first trying PhotoRec and it has found some files so i am not completely lost even though the PRO DUO said that the drive was empty!!!! But UFS explorer has the potential for recovering the directory structure so i'm checking that out. because going through thousands of files (what photorec and foremost find) with unknown names but known file extensions still isn't a pure victory but it's winning a mini battle.
I did get an error with the Standard recovery program from UFS. I was unable to copy it to the clipboard, it took 4 hours to compete a recovery scan (but the program died with a please send this to suppport), so i'll try to repeat the error tomorrow and post the error number.
Photorec works on linux. You can't directly mount these drives on linux. XFS is more of an SGI operating system but is supported by most linux forms. EXT2 or EXT3 is the normal linux filesystem. The filesystem on the PRO DUO is a bunch of silly primary partitions then an extended partition inside where the XFS partition sits. But that partition is labeled in a funny way so it won't be recognized by a non raid controler. You can peer into the partition itself with dd if=/dev/sdxx where xx is the partition drive and number. BUt you need a program to make sense of the data, especially when it's been initialized by a flash. I'm hoping UFS will do more than just locate the files but actually identify them as being in directories, so far it's not working smoothly.
Klaborde- In the case of a drive failure on a Linkstation Pro Duo, the first thing you'll want to try is performing a disk structure, which can be done by going into the web interface for the unit, then going to disk management, raid setup, then clicking on the raid array that contains the drive that is having trouble. Once in this page you should be prompted to check the problem disk and then hit restructure. This will give you an error indicating possible data loss, but since you have raid 1 set up the data will be mirrored on the other drive and you will not lose any. After the disk restructure completes it will rebuild the raid array, which may take a while.
If the restructure fails and problems persist with the drive, or if you are not prompted for a restructure, or cannot access the web interface for the unit, we will want to look into replacing the Linkstation for you. Originally we could replace the individual drive, but at this point due to certain regulations that davo referred to earlier we are not able to process replacement orders on individual drives from this unit, due to there being exposed wiring in the device, nor can we recommend you replacing the drive yourself, we can only recommend that we process a replacement order for the entire linkstation.
There have been several other helpful posters in this thread that have outlined the features and flaws of Raid technology, and while we do heartily endorse this technology as part of your data redundancy solution, as with any technology it is an excellent idea to be aware of Raid's shortcomings. The one that causes the most confusion and frustration amongst customers is that Raid only protects against drive failures. Errors that we see come up that can cause data loss/loss of access to the device include the Raid structure itself failing, the hardware other than the drive on the device becoming defective, and the firmware on the unit becoming corrupt. While these problems are rare, they are potential problems that anyone wanting to use the Raid technology should take into consideration when deciding on what methods to use to preserve their data. We want to assure you that we will do whatever is possible to help you regain access to the data stored on the unit should something like this happen.
Good points. Certainly RAID systems are only useful when drives fail, but drives should fail much more often than raid hardware. So having a device that has it's RAID hardware/software fail more likely than the drives is really not good business practice.
UPDATE on my failed raid1. I've submited the report to UFS explorer and they gave me a beta version to try,if this can recreate part of the directory structure, i will have resolved the failure. But even if not, i am still able to read files with PHOTOREC, they give them sequenced names so i'll have to figure out which file was which name but that is better than losing all the data.
Also, I don't know why people are saying there are "exposed wiring", the drives look like normal sata drives you can plug into any sata controler as i was able to do.
buscha wrote:
Photorec works on linux. Youcan't directly mount these drives on linux. XFS is more of an SGIoperating system but is supported by most linux forms. EXT2 or EXT3 isthe normal linux filesystem. The filesystem on the PRO DUO is a bunchof silly primary partitions then an extended partition inside where theXFS partition sits. But that partition is labeled in a funny way so itwon't be recognized by a non raid controler. You can peer into thepartition itself with dd if=/dev/sdxx where xx is the partition driveand number. BUt you need a program to make sense of the data,especially when it's been initialized by a flash. I'm hoping UFS willdo more than just locate the files but actually identify them as beingin directories, so far it's not working smoothly.
ddis not really necessary. If the drive is part of a RAID 1 array, thereis no striping or anything strange that would prevent a linux box fromreading it. The only thing that would prevent you from reading it onLinux with XFS support installed would be the filesystem label itself,as you mentioned. There is an easy workaround, however. First run"fdisk -l" to list the partitions, look for the largest partition onthe drive. Then run "file -s /dev/sdXX", this will tell you if thefilesystem is XFS for sure.
root@linuxmachine:~# mount /dev/sdb4 /mnt/data
mount: unknown filesystem type 'linux_raid_member'
root@linuxmachine:~# mount -t xfs /dev/sdb4 /mnt/data
root@linuxmachine:~# ls /mnt/data
drwxrwxrwx 2 root root 22 Feb 13 18:43 FTPTest1
drwxrwxrwx 2 root root 6 Jan 15 02:39 FTPTest2
drwxrwxrwx 12 root root 4096 Mar 18 14:36 Share
drwxrwxrwx 2 root root 6 Mar 13 11:52 restricttest
"file -s /dev/sdXX" will tell you info about the contents of the partition:
root@linuxmachine:~# file -s /dev/sdb6
/dev/sdb6: SGI XFS filesystem data (blksz=3D4096, inosz=3D256, v2 dirs)
Obviously this won't tell you if the data's intact, only if there is a filesystem on the volume.
I have used ufs explorer and photorec to get all that i could from my two raid1 drives that my pro duo now sais are in raid0 mode. I decided to try xfs_repair. In -n mode, xfs_repair says it finds a secondary super block that it can recover from so i've got not much to loose. I tried it to my second drive. It did a lot of stuff, and found a lot of errors. But finished successfuly. But i still can't mount the drive even when i specify xfs.
[root@localhost ~]# file -s /dev/sdd6
/dev/sdd6: SGI XFS filesystem data (blksz 4096, inosz 256, v2 dirs)
[root@localhost ~]# ls /mnt/x
[root@localhost ~]# mount /dev/sdd6 /mnt/x
mount: unknown filesystem type 'mdraid'
[root@localhost ~]# mount -t xfs /dev/sdd6 /mnt/x
mount: /dev/sdd6 already mounted or /mnt/x busy
[root@localhost ~]# xfs_repair /dev/sdd6
xfs_repair: warning - cannot set blocksize on block device /dev/sdd6: Device or resource busy
The drive is not busy (mounted) nor is my mount point. I'm not sure why xfs_repair appears to be able to change the superblock but it can't set the block size or be mounted. I know the data is still on there because i've recovered most of it. But am i trying to get the filestructure back. UFS explorer was probably the most promising product, but in the end it only was able to recover 1% of the files i could recover from Photorec, Foremost came in second by recovering only a few files that photo rec didn't already have. Since the files are unnamed it is still an option try to recover some of the directory structure so i can get to the files i really need that haven't been recovered yet by directory. Still 10% hopefull.
I always wondered what was the procedure for replacing a bad drive in my LS Duo Pro. There was certainly no documentation on it, which was amazing in itself, and an indicator of how much confidence Buffalo has in this product's ability to recover from this scenario.
So I was happy to find this topic, which I eagerly read through in its entirety, looking for something remotely reassuring. OMG...didn't find a shred of anything that inspired the remotest degree of confidence.
I have wasted my money, and half my NAS drive space on a completely useless implementation of "RAID".
I don't have a Linux system, and don't care to buy one just so I can try to diddle around with this piece of poorly-designed junk. I wonder what percentage of Buffalo users are unix people who are a) qualified, and b) willing, to do this?
This has become an expensive boat anchor. No, actually, it's always been that way, I just now figured it out, so shame on me.
This is the last Buffalo product that will ever enter my house.
I agree this is our first and last buffalo purchase. As soon as we can get the funds were replacing this unit with something from Iomega. Might be a little more expensive but as they say you get what you pay for.
dieseljwt wrote:
I always wondered what was the procedure for replacing a bad drive in my LS Duo Pro. There was certainly no documentation on it, which was amazing in itself, and an indicator of how much confidence Buffalo has in this product's ability to recover from this scenario.
So I was happy to find this topic, which I eagerly read through in its entirety, looking for something remotely reassuring. OMG...didn't find a shred of anything that inspired the remotest degree of confidence.
I have wasted my money, and half my NAS drive space on a completely useless implementation of "RAID".
I don't have a Linux system, and don't care to buy one just so I can try to diddle around with this piece of poorly-designed junk. I wonder what percentage of Buffalo users are unix people who are a) qualified, and b) willing, to do this?
This has become an expensive boat anchor. No, actually, it's always been that way, I just now figured it out, so shame on me.
This is the last Buffalo product that will ever enter my house.
Hmm you do realise that linux is free right? And also you say that you are not willing to do this, so what exactly ARE you willing to do? Power it on maybe?
Have you actually tried replacing the ******* disk or are you just going to moan and groan about every little hiccup you come across?!
Buffalo say they do not replace the disks so why the hell would they document how to do so?
You don't even say if you still have access to your data? Do you???!!! If so then the RAID has worked correctly so what are you complaining about?
If you do not have access to this then how do you know its not the controller or the RAID itself?!!!
Sorry to go on like this but i am sick to death of people on this forum coming on and ranting and raving about something that the vast majority (not all) don't really understand!!
If you would just take the time to explaing the situation then there are people here that give up there time (FOR FREE I MIGHT ADD) to help you come to some kind of conclusion!
P.S I really don't think it would make a very good boat anchor :P
Im sorry but when the makers of this drive tell you it should be backed up, whats the point? Thats what most people bought it for to begin with. If thats to simple a concept to grasp Im sorry. I didnt have to buy this drive to get a cheap software 'raid' I could have done that on my own.
So let me understand your logic, We should buy a buffalo to back up our files. Thats one backup. Then we should buy something to back up the buffalo. Thats the second backup. Well our company calls that money wasted.
We prefer to backup our files on one system that functions properly. If you cant grasp that, its your problem.
Actually its called disaster planning, you obviously have never heard of an offsite backup?! Maybe you should look into it ;-)
Fail to prepare, prepare to fail.
Yes a second device, if you read carefully here which you havent you would need a third device. Which defeats the purpose of buying the second one.
If your data is that important then there should be no arguments about this. Common good practice is to have multiple backups.
To respond to dieseljwt:
If you are using the Buffalo for a backup, it is doing it's purpose. Since you already have two copies (one on your local drive and one on the buffalo.) If there is a catastrophe, the recovery process is not that difficult: do download of an autoboot linux dvd which can temporarily turn your windows computer into a linux machine and then you'd be able to run xfs_repair no problem (for free.) After recovering my data and reformating back into mirror mode, i've been able to successfully use my buffalo as my backup system with out failure. No data was lost in my catastrophe after i repaied, in fact no data was lost on either mirrored drive. There was an error on the controller which cleared the boot sectors of the drives, which is fixable. I actually see no way to fix any of the ntfs drives that other NAS's provide, if there is any sort of error in an NTFS partition, you have to give up. Buffalo uses XFS which is a much simpler and faster filesystem than NTFS and thus more recoverable when things go wrong.
According to the manual, and from what i have always understood of RAID1 , when a drive goes bad, you replace that drive, and then when you power on the device it should automgically start rebuilding the info (Mirroring) on the new drive.
Exactly!!!!
I would suggest reading up first on what RAID is exactly and how to make yourself a suitable backup plan.
RAID 1 mirrors the contents of the disks, making a form of 1:1 ratio realtime backup. The contents of each disk in the array are identical to that of every other disk in the array. A RAID 1 array requires a minimum of two drives. RAID 1 mirrors, though during the writing process copy the data identically to both drives, would not be suitable as a permanent backup solution, as RAID technology by design allows for certain failures to take place.
Please note:
RAID 1 mirrors, though during the writing process copy the data identically to both drives, would not be suitable as a permanent backup solution, as RAID technology by design allows for certain failures to take place.