Skip to content
View in the app

A better way to browse. Learn more.

Unraid

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

A few questions about unRAID and whether or not it's for me...

Featured Replies

Alright, so I've been giving unRAID a look and it definitely looks promising as a future storage solution for my needs.

 

Here's my situation; I have an Adaptec RAID 31605 and I've enjoyed my time in RAID5, but the lengthy process for an online capacity expansion has me thinking whether or not RAID 5 is really appropriate for my needs. I like the performance striped parity gave me, but Adaptec products weren't exactly Arecas in the realm of sequential read/write speeds (I usually capped around 250MB/s or so but in real world performance, I would be lucky to get 150MB+ to be honest).

 

And now, I got myself a Norco RPC-3116 rackmount server, and for some reason when I booted up my system, my RAID card "yelled" (a loud, continuous beep) signifying that a drive had "failed". My array of 5 ST31500341AS (not the most popular drive out there, I know...) has disappeared and I am unable to access my data. Essentially my data is in a state of limbo and I will need to purchase temporary storage if I want my data back in order to create a temporary RAID array in order to retrieve my data.

 

This is where unRAID comes in - the more I read about its features the more and more I find that I like it; the ease of capacity expansion, compartmentalized damage if more than one disk were the fail, and so on. Essentially, I would be trading the RAID5 performance that I don't really need for all of these features, and to me, that trade off is worth it. So essentially, my RAID card will just become a gigantic 16-port SATA card, but I'm content if I can simply have massive storage without much hassle or lengthy times for online capacity expansions in the future.

 

Now, here's where I'm having a few doubts: I'm mainly a Windows user, meaning that I don't really have much Linux experience and whatnot, and I've enjoyed my server for other things not just for storage, but for a personal game server hoster, FTP server, TeamSpeak Server, and other various tasks. This is what I'm hoping I can do to best fit my needs:

 

My Setup:

Intel core 2 Duo E6400

2 x 2GB Corsair XMS 2 DDR2-800 RAM

6 x ST31500341AS (5 x SD1A FW, 1 x CC1G FW; firmware shouldn't matter as it won't be intensively used as in a RAID 5)

Norco RPC 3116

1 x WD2500KS (my Server 2008 OS drive)

1 x thumb drive for the unRAID setup

 

1. Create an unRAID Server Pro flash stick (I want the 16-drive support) and eventually expand the array as I need the storage...potentially getting WD20EADS as I need 'em.

 

2. Run the unRAID setup headless; have my actual Windows Server 2008 OS disk see the unRAID array as a large logical storage pool (essentially, a large disk) formatted in NTFS as that's the file system that is compatible with my setups in my home, and have the capability to read/write/format/delete/etc at will.

3. Maintain the unRAID array through the server interface as necessary, but it should largely go smoothly and reliably.

 

Is what I'm asking for too much? I'm really tired of my RAID 5 shenanigans to be honest...it's not very flexible and really is for those who have large budgets and require large I/O processes, which I don't necessarily need. I just want flexible, redundant storage that doesn't have much overhead (WHS is largely out because I want to duplicate all my data, and a gigantic RAID 1 overhead is just too much for me). If I can run the unRAID logical storage pool and have Windows Server 2008 see the array as an unformatted pool, I will be sold on the idea for unRAID for my next setup!

... I have an Adaptec RAID 31605 ...

 

I'm not sure if you'll find support for this card in unRAID.  I see no reference to it.  The good news is that it is really easy to test and see if it is supported with the free version of unRAID.

 

If the card does not work, you can request that support be added.  Tom (author) has been very responsive to requests for new disk controller support. 

 

.. when I booted up my system, my RAID card "yelled" (a loud, continuous beep) signifying that a drive had "failed". My array of 5 ST31500341AS (not the most popular drive out there, I know...) has disappeared and I am unable to access my data. Essentially my data is in a state of limbo and I will need to purchase temporary storage if I want my data back in order to create a temporary RAID array in order to retrieve my data.

 

Ouch!  Sorry to hear. This is definitely one of the advantages of unRAID - if a disk goes south you can rebuild it - and even if 2 disks were to go bad, the other disks are not affected AND you are still left with access to disk-level tools to try to recover data.  You do give up high performance, esp. on writes, but you do gain quite a lot of reliability.

 

... Now, here's where I'm having a few doubts: I'm mainly a Windows user, meaning that I don't really have much Linux experience and whatnot, and I've enjoyed my server for other things not just for storage, but for a personal game server hoster, FTP server, TeamSpeak Server, and other various tasks.

 

Lack of Linux knowledge is not an issue.  unRAID is an appliance.  There is no need to install an OS or anything.  Most users can do the install in way less than an hour.

 

You will not be able to run Windows apps on your unRAID server.  There are a number of tools that you can add to your unRAID server (look in this thread). If you want to run Windows WHS, it will be on a different server.

 

... If I can run the unRAID logical storage pool and have Windows Server 2008 see the array as an unformatted pool, I will be sold on the idea for unRAID for my next setup!

 

You lost me here.  unRAID storage will be available as networked storage space - not local storage space.  I don't think what you are suggesting here is possible.

 

  • Author

Right.

 

I appreciate your help, and what I meant by my last statement, I suppose that wouldn't really make any sense...however, is there a way so that I can setup the unRAID storage pool (even if it is seen as network storage) and run my Windows Server 2008 OS with the ability to read/write/delete/partition the storage pool from the server itself? So, in other words, I simply set up the unRAID storage pool from the flash stick, and then once that's done, I boot into my Windows Server 2008 disk and just run things from there?

 

What I'm concerned about is the NTFS file system and its ability to deal with disk failures within what is essentially a JBOD...would that be an issue, or does unRAID use its own file system?

You could certainly access the unRAID shares from the WHS server, and run things from there and store things to there.  You could even map a drive letter if you wanted.  It's just that it won't be part of a WHS drive pool.  Also, unRAID does not use that term (drive pool) to refer to its storage, so your terminology is a little off.

The unRAID software is a slimmed-down Linux distribution, that uses one of the common Linux file systems, ReiserFS.  Other file systems (such as NTFS) can be read (with a little work), but currently cannot be a part of an unRAID array.

 

However, that does not matter across the network.  Network storage is not accessed as a file system, but through 'protocols' like SAMBA, SMB, NFS, etc.  Windows Server 2008 will not be able to directly control the array, as an unformatted pool, but can fully access it through SAMBA and possibly NFS too (don't know for sure on that, but someone else here will know).  The unRAID server will look like any NAS to Windows Server 2008.

  • Author

The unRAID software is a slimmed-down Linux distribution, that uses one of the common Linux file systems, ReiserFS.  Other file systems (such as NTFS) can be read (with a little work), but currently cannot be a part of an unRAID array.

 

However, that does not matter across the network.  Network storage is not accessed as a file system, but through 'protocols' like SAMBA, SMB, NFS, etc.  Windows Server 2008 will not be able to directly control the array, as an unformatted pool, but can fully access it through SAMBA and possibly NFS too (don't know for sure on that, but someone else here will know).  The unRAID server will look like any NAS to Windows Server 2008.

Alright, it's looking encouraging thus far.  ;D

 

This is what I was thinking of doing:

 

1. Initially set up my unRAID disk setup as one large logical storage pool through User Shares.

 

2. Then, boot into my Windows Server 2008 OS disk (WD2500KS) that resides within my rackmount; this will be my primary operating system which will run various other serving tasks while also hosting my unRAID storage pool (which Windows Server will hopefully see as a network disk, am I correct)? Run various serving programs that utilize this storage pool (FTP, VPN, streaming, etc) and be on my merry way.

 

Is this a correct assumption that can be done with unRAID? I'm definitely gonna try seeing if simple volumes set via my RAID card will allow unRAID to detect the disks as individual drives, but looking from other compatibility lists, looks like some lower-end Adaptec non-RAID cards are supported, which I hope will translate into my Adaptec 31605 controller card as well.

 

I appreciate your help in all of this, and I hope to get more information about unRAID as time goes by, but so far, I'm quite impressed.  :)

  • Author

One more question, I've probably stated this in another form previously, but I really need to know this:

 

Can unRAID be run headlessly? As in, I boot into the flash drive, set up the array whenever needed (initial configuration of user shares and such or the expand/maintain the array), shut down, remove the flash drive, and then boot into my Windows Server 2008 OS disk that's within the server and be able to see the unRAID array from within the Server OS as a network disk? Or is this impossible without having the unRAID server running from the flash disk, which means I won't be able to use any Windows Server 2008 functionality?

One more question, I've probably stated this in another form previously, but I really need to know this:

 

Can unRAID be run headlessly? As in, I boot into the flash drive, set up the array whenever needed (initial configuration of user shares and such or the expand/maintain the array), shut down, remove the flash drive, and then boot into my Windows Server 2008 OS disk that's within the server and be able to see the unRAID array from within the Server OS as a network disk? Or is this impossible without having the unRAID server running from the flash disk, which means I won't be able to use any Windows Server 2008 functionality?

 

I am fairly certain you WILL NOT be able to do this.  unRAID has to be running for it to work.  I will not (if i am understanding your question correctly) be able to setup the array in unRAID and then boot into Windows Server 2008 and it still be "there."  The only way you might get this to work is by running VMWare on the Server 2008 Machine and then setting up the unRAID array in the virtual machine.

One more question, I've probably stated this in another form previously, but I really need to know this:

 

Can unRAID be run headlessly? As in, I boot into the flash drive, set up the array whenever needed (initial configuration of user shares and such or the expand/maintain the array), shut down, remove the flash drive, and then boot into my Windows Server 2008 OS disk that's within the server and be able to see the unRAID array from within the Server OS as a network disk? Or is this impossible without having the unRAID server running from the flash disk, which means I won't be able to use any Windows Server 2008 functionality?

 

OMG - that would be some trick!!!   ;D ;D ;D

 

Running headless means running without a video card (and therefore no monitor) attached.  unRAID can run headless with certain motherboards.

 

Running unRAID by osmosis is not going to work, on any motherboard!  If the unRAID OS ain't running, the array is not there.  And there is no way to run two OSes on the same machine at the same time without virtualization (see below)

 

What you are suggesting is similar to asking - if I put a music CD into my DVD player, play a song, can I then take the CD out, put in a DVD, and have the CD music play with the movie?  It is the same answer - no!

 

Trying to get a VM solution running is also not viable IMO.  There have been some attempts to run unRAID in a VM, but it is very limited.

 

Bottom line - if you want to run unRAID and WHS you need TWO machines.

 

  • Author

Bottom line - if you want to run unRAID and WHS you need TWO machines.

 

I hear ya!

 

This still doesn't really dissuade me from unRAID concept, I am willing to sacrifice the performance of RAID 5 striping for a no-headache, easy-to-expand storage solution, despite losing my old OS functionality (I guess I'll just run my desktop to do those functions over-the-network, it's not a biggie i suppose).

 

However, I've run into another potential brick wall when considering my unRAID solution: my Adaptec 31605 is probably incompatible with unRAID's current kernel; I have no Linux experience as I said, and I don't really have the time to try to integrate Adaptec's linux drivers into the current version of unRAID, so these are my current possibilities:

 

1. Try to do it myself (probably end in frustration and crash-and-burn results!  :o)

 

2. Hopefully a kind soul here will be able to do it for me and others.

 

3. Ditch the Adaptec controller (hopefully sell it for a lot) and just use simple motherboard SATA ports, PCI/PCIe 1x SATA controller cards (Silicon Image-based cards) and run it from there.

 

What I'm curious about is the performance difference between the limited thoroughput of the PCI bus versus the IOP processor within the Adaptec Card at handling unRAID I/Os; I understand there is no striping, but having it all within one large and very capable controller card would certainly be faster than multiple cards on multiple buses? What sort of performance would I expect to get with such a setup if I were to just simply try to connect 16 drives throughout the motherboard and external controller cards?

  • Author

Also, another quick question: once the unRAID server OS is booted up and the unRAID array is set up, can the USB flash stick be removed? I don't know why, but the idea that the entire configuration of the array is on one flash stick that can be lost so easily is eerily unnerving...I guess I'd have to make regular backups of the configuration files inside the flash drive!  :o

Also, another quick question: once the unRAID server OS is booted up and the unRAID array is set up, can the USB flash stick be removed? I don't know why, but the idea that the entire configuration of the array is on one flash stick that can be lost so easily is eerily unnerving...I guess I'd have to make regular backups of the configuration files inside the flash drive!  :o

Short answer is no.  It needs to be left plugged in.

See the wiki FAQ answer here:

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

 

If you make a copy of the "config" folder on the flash drive, and store it elsewhere on a different disk, you will be able to set everything back if you ever switch to a different flash drive.  This only need to be done after you make configuration changes (to assigned drives, shares, users, permissions, etc)

 

Joe L.

Also, another quick question: once the unRAID server OS is booted up and the unRAID array is set up, can the USB flash stick be removed? I don't know why, but the idea that the entire configuration of the array is on one flash stick that can be lost so easily is eerily unnerving...I guess I'd have to make regular backups of the configuration files inside the flash drive!  :o

 

"The entire configuration" consists of about 5 or 6 files that take up < 10K!  It can easily be backed up over the network.

 

If you are concerned about someone with physical access to the server grabbing the the USB and putting it in their pocket, you can mount the USB drive inside the case. This can easily be done by routing a USB extension cable into the case. A better option is to use an internal header and mount the USB completely insdie the case.

From a risk management POV if your concerned about the USB drive being taken you have bigger problems than that to worry about

Archived

This topic is now archived and is closed to further replies.

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.