1. Does the parity and cache drive count as one of the 20 max drives?

2. If your motherboard, server case, or multiple hotswap bays fail (but DATA is still there) do you lose the data on those drives?

3a. Does unRAID keep single files on one hard drive, or does it split it up between hard drives (like Raid 0/5)?

3b. If unRAID splits files up between multiple hard drives, wouldn't that corrupt data across all hard drives in the system if you had a 2 drive failure?

5. Can you easily swap out hardware (Motherboard, Ram, CPU, etc), even with different models, without rebuilding the array?

6. Can you defrag your hard drives with unRAID?

7. If you have a problem and unRAID isn't reading your hard drive correctly (for whatever reason), but the data is still there, can you just plug it into another computer and get the data off of it?

8. What file system does unRAID use?

 

Thanks to everyone that helps me, I know its a lot of questions. I might reply with more questions in I can think of any.

1. Does the parity and cache drive count as one of the 20 max drives?

2. If your motherboard, server case, or multiple hotswap bays fail (but DATA is still there) do you lose the data on those drives?

3a. Does unRAID keep single files on one hard drive, or does it split it up between hard drives (like Raid 0/5)?

3b. If unRAID splits files up between multiple hard drives, wouldn't that corrupt data across all hard drives in the system if you had a 2 drive failure?

5. Can you easily swap out hardware (Motherboard, Ram, CPU, etc), even with different models, without rebuilding the array?

6. Can you defrag your hard drives with unRAID?

7. If you have a problem and unRAID isn't reading your hard drive correctly (for whatever reason), but the data is still there, can you just plug it into another computer and get the data off of it?

8. What file system does unRAID use?

 

Thanks to everyone that helps me, I know its a lot of questions. I might reply with more questions in I can think of any.

 

Don't take this the wrong way but all of these questions have been asked before and cane be found by using the search function of the forum, via the unRAID wiki, FAQ, Best of the Forums, and the Topical Index.  Please do some searching and reading there to find the answer; if you can't find them after that come on back and we can point you in the right direction.

1. Does the parity and cache drive count as one of the 20 max drives?

2. If your motherboard, server case, or multiple hotswap bays fail (but DATA is still there) do you lose the data on those drives?

3a. Does unRAID keep single files on one hard drive, or does it split it up between hard drives (like Raid 0/5)?

3b. If unRAID splits files up between multiple hard drives, wouldn't that corrupt data across all hard drives in the system if you had a 2 drive failure?

5. Can you easily swap out hardware (Motherboard, Ram, CPU, etc), even with different models, without rebuilding the array?

6. Can you defrag your hard drives with unRAID?

7. If you have a problem and unRAID isn't reading your hard drive correctly (for whatever reason), but the data is still there, can you just plug it into another computer and get the data off of it?

8. What file system does unRAID use?

 

Thanks to everyone that helps me, I know its a lot of questions. I might reply with more questions in I can think of any.

 

Don't take this the wrong way but all of these questions have been asked before and cane be found by using the search function of the forum, via the unRAID wiki, FAQ, Best of the Forums, and the Topical Index.  Please do some searching and reading there to find the answer; if you can't find them after that come on back and we can point you in the right direction.

 

I have been through the wiki, and have read a lot about unRAID over the last week, and I still have not found answers to these questions.

 

No offense, but these are mostly yes or no questions, I don't see what the big deal is. Takes someone 2 minutes to answer my questions... Takes me hours to search around some more and maybe find the answers.

1. Does the parity and cache drive count as one of the 20 max drives?

http://lime-technology.com/wiki/index.php?title=FAQ#What_is_the_difference_between_the_unRAID_licenses.3F

 

2. If your motherboard, server case, or multiple hotswap bays fail (but DATA is still there) do you lose the data on those drives?

no, but the hotswap bays can cause problems if they take drives offline while data is being moved.  In general it is easier to not deal with the hotswap bays (and unRAID does not like when you try to hotswap a drive... at all)

 

3a. Does unRAID keep single files on one hard drive, or does it split it up between hard drives (like Raid 0/5)?

on one drive

 

3b. If unRAID splits files up between multiple hard drives, wouldn't that corrupt data across all hard drives in the system if you had a 2 drive failure?

see above

 

5. Can you easily swap out hardware (Motherboard, Ram, CPU, etc), even with different models, without rebuilding the array?

yes

 

6. Can you defrag your hard drives with unRAID?

no need to if you are using it as a file server.  with use of the cache drive the files get written in a very sequential order

 

7. If you have a problem and unRAID isn't reading your hard drive correctly (for whatever reason), but the data is still there, can you just plug it into another computer and get the data off of it?

yes, via windows with a driver that allows you to read the Reiser FS or on pretty much any linux distro

 

8. What file system does unRAID use?

Reiser FS

 

Thanks to everyone that helps me, I know its a lot of questions. I might reply with more questions in I can think of any.

 

 

Sorry for the turs  reply before.  The above should be the answer you are looking for, but the FAQ section of the wiki does answer a lot of these.

1. Does the parity and cache drive count as one of the 20 max drives?

http://lime-technology.com/wiki/index.php?title=FAQ#What_is_the_difference_between_the_unRAID_licenses.3F

 

2. If your motherboard, server case, or multiple hotswap bays fail (but DATA is still there) do you lose the data on those drives?

no, but the hotswap bays can cause problems if they take drives offline while data is being moved.  In general it is easier to not deal with the hotswap bays (and unRAID does not like when you try to hotswap a drive... at all)

 

3a. Does unRAID keep single files on one hard drive, or does it split it up between hard drives (like Raid 0/5)?

on one drive

 

3b. If unRAID splits files up between multiple hard drives, wouldn't that corrupt data across all hard drives in the system if you had a 2 drive failure?

see above

 

5. Can you easily swap out hardware (Motherboard, Ram, CPU, etc), even with different models, without rebuilding the array?

yes

 

6. Can you defrag your hard drives with unRAID?

no need to if you are using it as a file server.  with use of the cache drive the files get written in a very sequential order

 

7. If you have a problem and unRAID isn't reading your hard drive correctly (for whatever reason), but the data is still there, can you just plug it into another computer and get the data off of it?

yes, via windows with a driver that allows you to read the Reiser FS or on pretty much any linux distro

 

8. What file system does unRAID use?

Reiser FS

 

Thanks to everyone that helps me, I know its a lot of questions. I might reply with more questions in I can think of any.

 

 

Sorry for the turs  reply before.  The above should be the answer you are looking for, but the FAQ section of the wiki does answer a lot of these.

 

Thank you very much.

Actually my friend has one more question. We are thinking about a 2-pack of PRO.

 

Does unRAID work with external hard drives? (My guess is no)

Actually my friend has one more question. We are thinking about a 2-pack of PRO.

 

Does unRAID work with external hard drives? (My guess is no)

If they are eSATA, yes.  If just USB, no.

 

But... you cannot just remove a drive from the array without having to re-compute parity.  So, basically, once plugged in, they are part of the array and must remain plugged in.

 

Joe L.

...or replaced by another drive installed inside the machine.  UnRaid isn't a fan of reducing the number of disks in an array.

Actually my friend has one more question. We are thinking about a 2-pack of PRO.

 

Does unRAID work with external hard drives? (My guess is no)

 

well there is also this:

 

http://lime-technology.com/wiki/index.php?title=Mounting_an_external_USB_drive_having_an_existing_NTFS_file_system_in_READ/WRITE_mode_to_transport_files_from/to_unRaid_server

 

probably not what you were talking about though.

 

6. Can you defrag your hard drives with unRAID?

no need to if you are using it as a file server.  with use of the cache drive the files get written in a very sequential order

 

 

Okay this question gets asked all the time, and I mean absolutley no offense to anyone who has answered - but, I think the answer(s) are inadequate.

 

Why is it that "if it's used as a file server" unRAID should not need defragmenting?

 

If I may complain with an allusion to another industry: "the stats say the average person is 5 feet 9 inches tall so our headrests only have to extend that far up to satisfy government safety regulations" is hogwash!  Okay, if you're seven-foot-seven, I can see manufacrurers wincing at accommodating by default, but how many of us are 6-foot-3 and therefore not safe in our Chevy S-10 trucks?  (I don't drive one partly for this reason - I'm just saying)

 

Now, for most folks, sure maybe this "file server - so not needed" attitude suffices.  But I think we need a way - an easy way - to actually defrag unRAID for those of us who may be outside the normal parameters.

 

I do not know of one, other than perhaps moving files from fragged drives to newly-added (and empty) drives using the command line or MC.

 

Maybe it's a different issue I am having, but I am not sure.  I have 10 drives and do not use a cache drive.  I also take approximately 10 000 digital photos per year, and store all of them amongst the ten drives in my array.  Along with them are my other personal files, my CD collection in lossless format, and my DVD and BD collections in "lossless" formats as well.

 

Here's the scenario: With whatever I had lying around, 10 drives equated to roughly 6 TB.  After many more movie discs that was filled.  I bought larger drives and replaced the smaller ones.  So far so good.  This happened several times to the point where I am running 15TB currently on mostly 2TB and 1.5TB drives. 

 

Along the way when money was tight for a couple of months, the digital photos were still rolling in, along with the odd DVD or BD as gifts, etc.  So space was full.  I shifted digital pix from drive to drive to free up enough contiguous space to store the new stuff as needed.

 

More than once that scenario has happened, and my shares are well and truly scattered across the ten drives now. 

 

So, to the real problem: Being 2TB and 1.5TB drives, when I try to add a file to one that is nearly full (say 1.6GB free but I am trying to copy a few digital photos and other small items under 100MB each) the array takes forever to seek to a "free" spot on the drive.  (I use the fill-up allocation method and all shares have access to all drives.)

 

Tonight as I try to copy any small files at all onto the array, Windows times out with various errors because the array cannot find a spot to put the files, even though there are no parity errors or bad sectors and there is plenty of free space available.

 

So I guess I am looking for suggestions on this one.  I'll get through it by placing the files on a drive that's not so full manually, but it takes the fun out of it if I have to do that all the time.

 

Any thoughts?

Byron

1.6G spread across 10 drives means the you have 160M free space per drive on average. You can only write files up to 160M in size because each file must fit on a single disk. You should have the min free space of the user share set to at least twice the size of the largest file you will write. This is not a disk fragmentation issue. You have run out of space.

As dgaschk highlighted, this sounds like an example of fragmentation of free space across user shares, rather than fragmentation of files.  Such fragmentation can be reduced by careful configuration of the file allocation algorithms to ensure files end up where you would like them, or (as I do) write directly to disk shares so that you can manually coordinate the location of files and free space. 

Agreed, it is mostly fragmentation of shares, but why should that stop me from being able to fill a disk due to errors such as "the network device is no longer available" ( a time-out error in my experience ) or "this file is in use" because unRAID hasn't yet been able to write it to a spot on the disk?

 

I know my message is long-winded by way of explanation, but I did not state that I had 1.6GB free across ten disks.  I have 1.6GB free on drive 1, 68MB free on drive 2, 1.4GB free on drive 3, 162GB free on drive 4, etc.  In total I have more than 1.36TB free at the moment.

 

Because I use the fill-up method, and because I am trying to write a file under 300MB in size (the file size minimum I set for my digital pics share) then unRAID always wants to direct those files to drive one first.

 

Even when I try to manually place the files on drive one, Windows times out with various errors depending upon how fervently I have been trying.

 

It's not a free space issue, and I did as you have suggested by assigning minimum free space allocations per share a long time ago when I realized how useful it is.  My issue is the long delay before unRAID is ready with the necessary free sectors to write files to, and the files never get written.

 

Now if I change my allocation method to high-water, I could write files without issue as I have a couple of drives with lare chunks of free space still available.  But I am sure you'll agree with my conclusion that doing this would only delay the inevitable, and I'd be right back to this jam when all drives became very close to full.

 

The basic issue is that most all programs used to copy files to the unRAID server will start by creating an initial empty file and proceed to append to it as the copy occurs.  The initial small file fits the disk initially chosen by unRAID based on your allocation method and rules.  The full size of the file does not therefore, at some time the copy fails. 

 

The windows PC has very few "error messages" and it uses a message that does not tell the real cause.  None really state the issue as windows thinks there is the 1.6Gig free.  it does not know you tried to copy a 300Meg file to a disk with only 68Meg free.

 

When your disks get that full YOU must take notice of the free space and either change the allocation method of copy using the "disk" shares so the files have the room they need.

 

setting the min-free on the specific user share will help, although with the smaller size files you are copying and the varying amount of space available it may not help as much at this time.  (Can't hurt to set it to 100 meg at this time though... since the field is in 1K blocks, that would be a value of

100000 )

 

Joe L.

 

 

I have always been a bit of a glutton for punishment when it comes to filling media to capacity.  When I used my C=128 system for so many years, I eventually wrote some code to gain more disk space by reallocating sectors in Track 18 (the directory track was wholly reserved for file system use) for use by files, after first checking that there was enough "headroom" for several more files to be added to the disk.

 

 

And so it is today that I want that 1.6GB that's supposedly available on my disk 1 to store 300 more digital photos... but Windows XP just isn't letting it happen.

 

I'm off to a telnet prompt and some command line folly to make it right

Since ReiserFS is a journaling filesystem, you should ensure you always have some free-space available should you decide to do file renames or moves.

Journaling is an interesting concept I paid no attention to in the past...

 

So if I understand it correctly, ResierFS needs to be able to have a bit of space for inode and/or journal expansion, true?  And this space is dynamically allocated from what's left as "free space" on the drive, correct?

 

And if those are true, then how many KB would one expect to need for such purposes, if free space itself becomes a dynamic concept when considering the moving of files?