unRAID fork(), concepts, ideas and discussion, Running unRAID on another distro.


Recommended Posts

Guys, heads up.

 

I've created a new thread dedicated to 'fork' development discussion. It'd be great to keep the discussion in this thread going to really work out the pro's and con's of various approaches but it would be great to keep this new thread clear for questions directly relating to the topic.

 

PLEASE REFRAIN FROM PERSONAL ATTACKS.

Link to comment
  • Replies 153
  • Created
  • Last Reply

Top Posters In This Topic

Guys, heads up.

 

I've created a new thread dedicated to 'fork' development discussion. It'd be great to keep the discussion in this thread going to really work out the pro's and con's of various approaches but it would be great to keep this new thread clear for questions directly relating to OP.

 

http://lime-technology.com/forum/index.php?topic=31072.0

 

 

Pity that didn't work out as planned. Some people can be quite rude (and I even had sympathy for the contra point of view before everything derailed so spectacularly).

 

Chin up though, this is a good idea and I'm looking forward to some/any outcomes.

 

Peter

 

Link to comment

I hope they put the thread back (minus all the attacks which should go to bilge)

 

The mods did not remove it, it was likely removed by ironic.  He did not move it to the bilge so it is gone, unless Tom brings it back from a backup.

 

think there was some valid/useful discussions in there before the attacks.

Agreed

Link to comment

I hope they put the thread back (minus all the attacks which should go to bilge)

 

think there was some valid/useful discussions in there before the attacks.

 

I'd like to discuss a fork of unRAID. Which Distro to use, adding dual / triple parity, changing from Rieser to BTRFS, Xen or KVM or both, using repositories for updates and package manager to install packages (instead of plugins), etc. Whether or not Tom will provide a separate Forum section on here, a place for Wikis, Documentation, etc. for the unRAID community to contribute.

 

However, 2 or 3 users took over the thread and start attacking me (I was told to go F$#^ myself), other users and even mods who wanted to discuss it. It didn't matter that Tom and the mods were engaging in the thread / discussing it on their own Forum.

 

As far as those couple of users are concerned... A Fork of unRAID is an abomination and pure evil. They seem to think it's mere existence or even discussing it will self destruct their unRAID Server even though they don't download and install it.

 

Makes it hard for people like Ironic, Ford, myself, others to want to FREELY invest our time, money, effort and energy into helping Tom and the unRAID community when being personally attack, criticized and ridiculed. Not to mention, those same users who are doing the attacking contribute NOTHING to the community like the rest of us have with ESXi, Xen, KVM, documention, guides, how toos, providing / updating KVM / Xen unRAID Kernels, assisting users, responding to questions / threads, solving problems users are having, answering PMs, remoting in users Servers and fixing issues or teaching them how to get unRAID working in Xen / KVM, etc.

Link to comment

I hope they put the thread back (minus all the attacks which should go to bilge)

 

think there was some valid/useful discussions in there before the attacks.

 

I'd like to discuss a fork of unRAID. Which Distro to use, adding dual / triple parity, changing from Rieser to BTRFS, Xen or KVM or both, using repositories for updates and package manager to install packages (instead of plugins), etc. Whether or not Tom will provide a separate Forum section on here, a place for Wikis, Documentation, etc. for the unRAID community to contribute.

 

However, 2 or 3 users took over the thread and start attacking me (I was told to go F$#^ myself), other users and even mods who wanted to discuss it. It didn't matter that Tom and the mods were engaging in the thread / discussing it on their own their own Forum.

 

As far as those couple of users are concerned... A Fork of unRAID is an abomination and pure evil. They seem to think it's mere existence or even discussing it will self destruct their unRAID Server even though they don't download and install it.

 

Makes it hard for people like Ironic, Ford, myself, others to want to FREELY invest our time, money, effort and energy into helping Tom and the unRAID community when being personally attack, criticized and ridiculed. Not to mention, those same users who are doing the attacking contribute NOTHING to the community like the rest of us have with ESXi, Xen, KVM, documention, guides, how toos, providing / updating KVM / Xen unRAID Kernels, assisting users, responding to questions / threads, solving problems users are having, answering PMs, remoting in users Servers and fixing issues or teaching them how to get unRAID working in Xen / KVM, etc.

 

All true.

 

However I can't take someone with that avatar seriously! HAHA

 

8)

Link to comment

Any chance a mod could set up or set ground rules on a thread to discuss implementation issues with bans handed out for going off topic? A second thread could be set up to discuss the merits of the idea itself and to discuss plans and ideologies ( a bit like the 'is payment justified' thread).

 

It would be great to have just a technical discussion that doesn't run aground with personalised attacks.

 

Peter

Link to comment

I did remove it.

 

A simple comment that you weren't going to participate in the discussion anymore due to the attacks would have perhaps been a more appropriate response than just deleting the entire thread.    "Taking my ball and leaving" is not a particularly mature response ... that's twice now that this has been done in these virtualization threads (remember ddeeds ?)

 

The thread had indeed deteriorated to some rather personal attacks ... but a note to the moderators or Tom to please intervene would have been a much better approach.

 

 

Link to comment

Grumpy, please just ignore this idiots.

 

I think this idea of taking unRaid to the next level is great, and very needed.

Tom is only one man and maybe unable to do this by himself.

I wish  I could help too. yours and ironic  guides are the best by far I have seen.

I have tried and accomplished more things using them  and in far less time , than I did by simply googling for help on each issue I came across.

 

 

As for all the dumb-asses posting BS all over this forum attacking your efforts on the basis of their perception that you try do steel unraid form Tom I say that they need to really read the posts/threads before coming to any conclusions, thus it would save them from coming up looking as total morons .

 

I have been reading and researching Linux file server setups and virtualization for the last 5 month (ever since I got my self the server from TAMs and found that using it as a VM host is an easy thing) frankly my understanding of things let me to a conclusion that for the most people it is  possible to recreate the system capable of what unraid  is ,except for couple of useful capabilities that unraid was designed to provide , but totally FREE using right linux distro is a reality.

 

how? well

unriad was developed to overcome the issues that regular  RAID and LVM had/have

such as : 

1. regular RAID must be planned and created at one time.

    resizing and updating Regular RAID array is not easy or some time not even possible without fully disassembling and reassembling of the array (thus need for take all the data out and put it back in afterwards )

 

2.regular RAID  must use the same size drives always. you put bigger drive, you loose  the space on it until all the drives are the same size.

 

3. LVM solve the issue by adding easy resizing capability of the managed volume

and adding support for different size disks in the group but no protection of any kind unless you RAID the underling disks and add that to LVM

 

 

unRaid does all this seamlessly and make it easy to some degree.

 

BUT

using a current distro of Ubuntu or OpenSuse

along with the BTRFS + SAMBA will do all that as well and more .

 

the only thing you loose in comparison to unraid (and that is important, but not for all people) is that

#1. BTRFS based array volume acts and feels like a regular RAID so loss of a disk can mean a total data loss on all array and not only data on that disk.

#2. all disk need to spin up on read as opposed to unraid when only the disk containing the data spins up

 

so I for am really looking forward to see how the unraid transformation go as I do want to use it as my main file server capable  of VM and other goodies .

 

 

 

Link to comment

In any large user community you're always going to have some vocal ones who resort to childish personal attacks when they disagree with another's point of view. It's the nature of non-face to face communication.

 

I just wanted to chime in and say that there are likely far more of us that are interested in what you are doing and appreciate your work than there are at the other end of the spectrum. Don't let them discourage you.

 

Also, I do have to agree with garycase here, it would have been better to have the mods handle the issue than to delete the thread, but I know how things can be in the heat of the moment.

Link to comment

how? well

unriad was developed to overcome the issues that regular  RAID and LVM had/have

such as : 

1. regular RAID must be planned and created at one time.

    resizing and updating Regular RAID array is not easy or some time not even possible without fully disassembling and reassembling of the array (thus need for take all the data out and put it back in afterwards )

 

2.regular RAID  must use the same size drives always. you put bigger drive, you loose  the space on it until all the drives are the same size.

 

3. LVM solve the issue by adding easy resizing capability of the managed volume

and adding support for different size disks in the group but no protection of any kind unless you RAID the underling disks and add that to LVM

 

 

unRaid does all this seamlessly and make it easy to some degree.

 

BUT

using a current distro of Ubuntu or OpenSuse

along with the BTRFS + SAMBA will do all that as well and more .

 

the only thing you loose in comparison to unraid (and that is important, but not for all people) is that

#1. BTRFS based array volume acts and feels like a regular RAID so loss of a disk can mean a total data loss on all array and not only data on that disk.

#2. all disk need to spin up on read as opposed to unraid when only the disk containing the data spins up

 

so I for am really looking forward to see how the unraid transformation go as I do want to use it as my main file server capable  of VM and other goodies .

 

like that - more or less the same I wanted to say - I would love to see unraid on a full fletched linux distri - because of the raid features it provides. VM support would be cool but the main purpose still should be - RAID - not the rest ;)

 

thumps up

Link to comment

In any large user community you're always going to have some vocal ones who resort to childish personal attacks when they disagree with another's point of view. It's the nature of non-face to face communication.

 

I just wanted to chime in and say that there are likely far more of us that are interested in what you are doing and appreciate your work than there are at the other end of the spectrum. Don't let them discourage you.

 

Also, I do have to agree with garycase here, it would have been better to have the mods handle the issue than to delete the thread, but I know how things can be in the heat of the moment.

Well said.

 

+1  ;)

Link to comment

I intervened where I felt appropriate. I sent out a few PMs, bilged a post and alerted Tom.

Tom's busy working on 64bit unRAID and other features.  Really we need to be civil about this.

 

As the ironic one pointed out 'Civility costs nothing'. ;)

 

Part of me thinks that a fork() of unRAID isn't the way to consider it but a full distro'ed unRAID is a worthwhile endeavor if all parties are involved.

People ran unRAID on a full slackware distro years ago.

 

What is learned from one camp will eventually be brought to the other camp.

We've seen this with all the virtualization threads. Today we have a forum dedicated to it.

 

Truth be told.

I'm getting more frustrated with commenting on threads only to have authors remove the whole thing.

It's making me not want to invest my time on these things anymore.

Link to comment

I did remove it.

 

A simple comment that you weren't going to participate in the discussion anymore due to the attacks would have perhaps been a more appropriate response than just deleting the entire thread.    "Taking my ball and leaving" is not a particularly mature response ... that's twice now that this has been done in these virtualization threads (remember ddeeds ?)

 

The thread had indeed deteriorated to some rather personal attacks ... but a note to the moderators or Tom to please intervene would have been a much better approach.

 

I did this, I tried locking the thread before it deteriorated and also tried reasoning. It was too much smoke with fire tbh, it was detracting from the entire purpose of the thread and now this one is going that way too. I thought my responses were mature and reasonable yet things were spiralling so I just reacted and deleted one thread.

 

So for clarity, I will reply or respond to no further questions about history.

 

Sent from my Nexus 5 using Tapatalk

 

Link to comment

Apologies to all for removing the thread. I will leave them in future I hadn't considered the effect of people investing time into them only to be deleted, it was selfish. You live and learn.

 

Any who, I am not discouraged and nor do I intend to be at loggerheads with the existing product.

 

Am working on it right now!

 

Sent from my Nexus 5 using Tapatalk

 

 

Link to comment

how? well

unriad was developed to overcome the issues that regular  RAID and LVM had/have

such as : 

1. regular RAID must be planned and created at one time.

    resizing and updating Regular RAID array is not easy or some time not even possible without fully disassembling and reassembling of the array (thus need for take all the data out and put it back in afterwards )

 

2.regular RAID  must use the same size drives always. you put bigger drive, you loose  the space on it until all the drives are the same size.

 

3. LVM solve the issue by adding easy resizing capability of the managed volume

and adding support for different size disks in the group but no protection of any kind unless you RAID the underling disks and add that to LVM

 

 

unRaid does all this seamlessly and make it easy to some degree.

 

BUT

using a current distro of Ubuntu or OpenSuse

along with the BTRFS + SAMBA will do all that as well and more .

 

the only thing you loose in comparison to unraid (and that is important, but not for all people) is that

#1. BTRFS based array volume acts and feels like a regular RAID so loss of a disk can mean a total data loss on all array and not only data on that disk.

#2. all disk need to spin up on read as opposed to unraid when only the disk containing the data spins up

 

so I for am really looking forward to see how the unraid transformation go as I do want to use it as my main file server capable  of VM and other goodies .

 

like that - more or less the same I wanted to say - I would love to see unraid on a full fletched linux distri - because of the raid features it provides. VM support would be cool but the main purpose still should be - RAID - not the rest ;)

 

thumps up

 

well it's just get to me when some people would go all "full attack mode"  and it's not even their war.

 

if Tom have a problem with what grumpy and ironic are doing  it should be him, not a few selected idiots, to battle and stop it.

 

 

 

 

 

Link to comment

like that - more or less the same I wanted to say - I would love to see unraid on a full fletched linux distri - because of the raid features it provides. VM support would be cool but the main purpose still should be - RAID - not the rest ;)

 

Due to the personal attacks and those 2 or 3 people questioning me, what possiblities an unRAID Fork can do. I will put together a thread full of screenshots so everyone will understand / see it.

 

If I have time when I get home from work (if not I will do this tomorrow)... I will post 15 or more screenshots of the following:

 

Note: All of this is using 100% of the unRAID Kernel Module and emhttp.

 

1. Running unRAID on CentOS 6.5 in 64-Bit

 

2. Running unRAID on the host (CentOS 6.5) with some drives and using other drives on my system I will run ANOTHER unRAID in a VM in Arch Linux also 64-Bit. 2 unRAIDs (different Distros) on 1 system, using different drives, 1 motherboard and no PCI SATA/RAID/SAS Controllers.

 

3. Passing through a nVidia 430GT video card to the unRAID VM (running on a totally different Linux Distro - Arch) and running XBMC with full Hardware Video Acceleration. 2 unRAIDs (Host and a VM) running at once with different hard drives and with XBMC running on both the Host unRAID and the VM unRAID.

 

Note: You can't passthrough a nVidia card in ESXi or Xen unless you have Quattro Series or hack your ROM. So this is another NEW feature that would be available.

 

5. Both using NFS4 and Samba 4.

 

6. Use either KVM or Xen for Host and VMs.

 

7. Crypt support.

 

8. LVM, Device Mapper, Thin Provisioning, etc. support.

 

9. Full Package Manager with screenshots of how easy it is to install XBMC, Sickbeard, etc. ONCE instead of the current Plugin Model.

 

10. Using VirtFS. (VMs can access the host unRAID "shares" / data at near drive level speeds. NFS / Samba aren't even in the ballpark when it comes to that speed).

 

11. I have it running headless but I will install a Desktop, so you can see it all running and the CPU Utilization... It's next to NOTHING (I will post screenshot of the unRAID running on the Host and the VM both playing 1080p with HD Audio from the other unRAID server).

 

12. All of this runs 30% FASTER than unRAID does now.

 

You can do all of the above and WHATEVER THE HELL YOU WANT in your Linux Distro of choice. I can provide ISOs for openSUSE, Arch, CentOS, Debian, etc.

 

I will also enable DUAL / TRIPLE Parity support and move the File System from Resier to BTRFS if the users want that too.

 

I am leaving the choice up to Tom if / how he wants to proceed and if he wants to facilitate having all of this here or in a separate fork elsewhere. All I can do is ask Tom privately and publicly for the world to see (which I have done). If those 2 or 3 people cannot see that... I don't know what I or anyone else can tell them to convince them of that.

Link to comment

like that - more or less the same I wanted to say - I would love to see unraid on a full fletched linux distri - because of the raid features it provides. VM support would be cool but the main purpose still should be - RAID - not the rest ;)

 

Due to the personal attacks and those 2 or 3 people questioning me, what possiblities an unRAID Fork can do. I will put together a thread full of screenshots so everyone will understand / see it.

 

If I have time when I get home from work (if not I will do this tomorrow)... I will post 15 or more screenshots of the following:

 

Note: All of this is using 100% of the unRAID Kernel Module and emhttp.

 

1. Running unRAID on CentOS 6.5 in 64-Bit

 

2. Running unRAID on the host (CentOS 6.5) with some drives and using other drives on my system I will run ANOTHER unRAID in a VM in Arch Linux also 64-Bit. 2 unRAIDs (different Distros) on 1 system, using different drives, 1 motherboard and no PCI SATA/RAID/SAS Controllers.

 

3. Passing through a nVidia 430GT video card to the unRAID VM (running on a totally different Linux Distro - Arch) and running XBMC with full Hardware Video Acceleration. 2 unRAIDs (Host and a VM) running at once with different hard drives and with XBMC running on both the Host unRAID and the VM unRAID.

 

Note: You can't passthrough a nVidia card in ESXi or Xen unless you have Quattro Series or hack your ROM. So this is another NEW feature that would be available.

 

5. Both using NFS4 and Samba 4.

 

6. Use either KVM or Xen for Host and VMs.

 

7. Crypt support.

 

8. LVM, Device Mapper, Thin Provisioning, etc. support.

 

9. Full Package Manager with screenshots of how easy it is to install XBMC, Sickbeard, etc. ONCE instead of the current Plugin Model.

 

10. Using VirtFS. (VMs can access the host unRAID "shares" / data at near drive level speeds. NFS / Samba aren't even in the ballpark when it comes to that speed).

 

11. I have it running headless but I will install a Desktop, so you can see it all running and the CPU Utilization... It's next to NOTHING (I will post screenshot of the unRAID running on the Host and the VM both playing 1080p with HD Audio from the other unRAID server).

 

12. All of this runs 30% FASTER than unRAID does now.

 

You can do all of the above and WHATEVER THE HELL YOU WANT in your Linux Distro of choice. I can provide ISOs for openSUSE, Arch, CentOS, Debian, etc.

 

I will also enable DUAL / TRIPLE Parity support and move the File System from Resier to BTRFS if the users want that too.

 

I am leaving the choice up to Tom if / how he wants to proceed and if he wants to facilitate having all of this here or in a separate fork elsewhere. If those 2 or 3 people cannot see that... I don't know what I or anyone else can tell them to convince them of that.

 

I want OpenSuse  :-)  gime,gime,gimie  pleeeeesseee, pretty please with a cherry on top ....

Link to comment

I did remove it.

 

A simple comment that you weren't going to participate in the discussion anymore due to the attacks would have perhaps been a more appropriate response than just deleting the entire thread.    "Taking my ball and leaving" is not a particularly mature response ... that's twice now that this has been done in these virtualization threads (remember ddeeds ?)

That's easy to say now, but (verbal) violence was escalating. Mods (I believe you are one) , or even Tom, should have anticipated and stopped it, but they didn't. And ironic locked the thread, only to have it unlocked by a mod (I think it was NAS). I'm sure NAS meant well, as did ironic,  but coming now to say it shouldn't have been deleted is, IMHO, a cheap shot.

 

Ironic, grumpy and others are doing this for free, they don't have to take any bs. They have taken certainly more than I would have. And bs is defined per their terms, there is no absolute value on this, this is one of the few perks of doing this for free.

 

Finally, if this forum is not the place to discuss this thing, which I believe is a valid argument (although I don't share it) it's up to tom and the mods to say so.

 

Sent from my GT-I9305 using Tapatalk

 

 

Link to comment

I'd like to discuss a fork of unRAID. Which Distro to use, adding dual / triple parity, changing from Rieser to BTRFS, Xen or KVM or both, using repositories for updates and package manager to install packages (instead of plugins), etc. Whether or not Tom will provide a separate Forum section on here, a place for Wikis, Documentation, etc. for the unRAID community to contribute.

I think the key point is in here and the message is buried under the welter of posts around how it will be done or the politics around it. Some posters seem to express themselves rather aggressively but it's not hard to understand the root cause of such a response (e.g. people don't like change, they perceive a threat to their investment in the product, they don't like the cut of ironic/grumpy's jib).

 

Things might be a little less heated if it were a bit more obvious *why* this is a good thing and not a threat. To my mind there are 2 major reasons to do/support this;

 

[*]make unraid a better (NAS/file server) product

[*]sweat your hardware by making the same server do more

 

The latter is pretty obvious as the use case already has a solution in the plugin setup. I don't use it to run anything other than a few scripts but it doesn't strike me as a particularly good solution to the problem, it is a solution nevertheless.

 

However there appears to be a whole class of users who simply do not care about this aspect at all, they just want a reliable, robust & performant file server. For them, the former use case is the key and I don't see a clear and unambiguous statement of what such people would get out of this *in terms an end user would understand*. There are a lot of posts about *how* it would be implemented but not why it should matter to them & why they should support it. I'm not saying such points don't exist (e.g. comments about n parity or perf improvements) but that they are generally secondary to how it will be done (e.g. use brtfs, use kernel x.y.z, use virtfs).

 

I think it would be v useful if someone were to list out some of the benefits that this can actually deliver, as well as any downsides that may exist, as well as pointing out how/why this isn't possible on unraid as is. Ideally there'd be a subforum dedicated to this subject, a sticky post or 2 on what it aims to deliver and then n threads around the different aspects of *how* to deliver those features as the current approach of 1 monolithic thread covering all aspects simply doesn't work.

Link to comment

Geeze, how hard is it to understand a simple concept?

 

If you want to post guides and work towards making it simple for someone to use unRAID in a full distribution then have at it. Include Tom if he's willing and you want to or simply don't. Obviously, there will be advantages to doing so for certain people. Notice I didn't write making a new full distribution including the unRAID part. That could be a possibility. I haven't seen an answer on that being allowed or not.

 

If you want to discuss a software fork, raise support money and talk about replacing emhttp then you should take it elsewhere. Fork discussions have no place here. But then I guess people don't understand what fork means. Maybe this will help.

 

http://en.wikipedia.org/wiki/Fork_%28software_development%29

 

"In software engineering, a project fork happens when developers take a copy of source code from one software package and start independent development on it, creating a distinct piece of software."

 

Since the ones posting about forking have no affiliation with LimeTech this means completely new 100% independent development. Starting with what can be used from unRAID (no emhttp) and developing something without any future involvement of LimeTech. Can you get it now?

 

Link to comment

If you want to discuss a software fork, raise support money and talk about replacing emhttp then you should take it elsewhere. Fork discussions have no place here. But then I guess people don't understand what fork means. Maybe this will help.

 

 

Let's leave telling them to go elsewhere up to Tom. We've made him aware of this.

Your point is duly noted. Please stop.

Link to comment

FYI, it wasn't me who asked for donations originally but was an idea floated by others. As a full time student it would have been daft for me to ignore people who wanted to give me money towards the costs to attempt to make unraid in a full distro a possibility.

 

I have tried to be as transparent as possible. I am not 'raising funds' for a sister company or competitor they are donations towards me, personally because people wanted too. Those people are incredibly kind and generous too, and really fired me up to repay their generosity with a return.

 

Some people will like it, some won't. As has been said a dozen times, if you don't like me or the cut of my jib then just leave my threads alone or contribute yourself.

 

Once again, as I have stated a dozen times, I WANT to work WITH tom not at loggerheads but cannot do so until he responds to my emails now can I?

 

Sent from my Nexus 5 using Tapatalk

 

 

Link to comment
Guest
This topic is now closed to further replies.