Which is more important to you, unRAID 5.0 features or 3TB drive support?

[neilt0]

Can't say 3TB support matters much to me.  Price point sucks in comparison to the 2TB. 

 

Not really -- when you take in to consideration the price of the port as well as the price of the drive; 1/3 more space means it costs you 1/3 less in "port cash".

 

I'd like 3TB support in something that works in a production array. There's no point having 3TB support in 5.0 if it's in Beta testing for a long time and can't be trusted with your data. There are not going to be many of us who can spare several 3TB drives for a test-only rig.

 

So, for me, the priority is:

 

* 3TB support in a stable release.

 

The MBR problem in the current 5.0 release rules it out for me.

[squirrellydw]

5.0 than add the extras like 3tb support. I want a stable os before anything.

[SSD]

Looking at the new wording in the poll questions I get that the intent is to encourage voters to vote for 3TB and don't really match up with the title of the post anymore.

 

How about:

Do you want support for 3TB now even if it means stopping other 5.0 development for an indefinite period of time?

 

I'd rather leave time out of it and focus on priorities.  Obviously unRAID development is not a continuous activity, and has to do with priorities, some of which have to do with demand, some have to do with LimeTech strategic direction on unRaid, and some that have nothing to do with unRAID.  Hopefully it is clear between option 1 and 2 that 2 pushes out the Plugin manager and the activie directory piece into the next release.

 

I agree with those saying the MBR check bug needs to be fixed.  I did say in option 1 that it would include bug fixes.  That is clearly an important one that is limiting use of 5.0b6a by advanced users who might otherwise be helping to find bugs and report them.

 

As far as getting 3T "released" soonest, I'm in favor of getting 5.0 to release status rather than a new 4.8 release.  If it was a priority, I don't think either would need to take long.  I think there is enough interest in the 3T feature that it rises to "must have" status, and that Tom should include this in the 5.0 release.  But if you think 3T support should not be a priority, suggest you vote the fourth option.  I will clarify the test on that option.

 

As far as the enhancement to GPT partitions, I think there is a general sense that this is a big enhancement on the technical side.  We all thought the AF enhancement (aligned vs unaligned partitions) was going to be a big one, but it was not.  We learned that parity is maintained based on the partition and not the disk, which means GPT doesn't affect parity.  Obviously only Tom knows for sure, but he has already reported laying all the groundwork to support the GPT partitions in the 5.0 betas.  It is really just the boot record that is different.  The biggest part could be actually laying the boot record and creating the GPT partition (which Linux supports doing), and making sure it is recognized as valid (and not detected as alien and clobbered as some MBRs are in the 6a beta).  There may be some complexity that I am not considering when rebuilding / upsizing an MBR disk to the GPT disk, but I doubt it would be a big problem.

 

Kizer - sorry man.  Didn't mean to beat you up. 

[kizer]

Its all good. Thats what this thread is about. Getting opinions and dismissing rumors. LOL

 

My biggest contribution to the thread is safety for our data and like stated Tom would never do anything to risk that. I just don't want the rush to even question that at any cost.

 

I think I'm going to see if I can find info on this GPT stuff because its got my intrest peaked.

Not the ultimate source of info, but it helped me understand some.

http://en.wikipedia.org/wiki/GUID_Partition_Table

[Rajahal]

As I understand it, the biggest hurdle that Tom needs to overcome to allow for 3 TB support is that of rebuilding an 'old' drive (with an MBR) with a 'new' drive (with a GPT).  Currently unRAID just rebuilds a drive bit for bit, so it really has no knowledge of the boot record.  In order to support GPT, unRAID will either have to first check the boot record of a drive and determine the correct way to rebuild it, or it will have to start rebuilding only the data partition instead of the entire drive.

 

Someone please correct me if I have this wrong.

[lionelhutz]

No, unRAID parity only supports/maintains the partition or unRAID only rebuilds the partition bit for bit. It does not protect the whole disk. When a disk is replaced, unRAID first partitions the disk and then uses the parity to reconstruct the contents onto the disk.

 

This is what made AF fairly easy. Tom just added support to create/move the partition on sector 64 and all else worked as is.

 

Peter

 

[Rajahal]

I see, thanks.  Interesting that the poll is almost perfectly split between 3 TB support and no 3 TB support at the moment...looks like we might not reach a clear consensus.

[lionelhutz]

looks like we might not reach a clear consensus.

 

How often do we reach a clear consensus here? 

[aiden]

I see, thanks.  Interesting that the poll is almost perfectly split between 3 TB support and no 3 TB support at the moment...looks like we might not reach a clear consensus.

Seeing as how 3 of the 4 options are in favor of 3TB support in some way, I would not call it a perfect split.  Also, while there are NUMEROUS advances needed/wanted in the 5.x series, they are not as time sensitive as adding >2.1TB drive support.  AFP is only important to Mac users, AD is only important to business users.  >2.1TB support will affect most unRAID users as time rolls on.   As drive densities continue to increase, the need for 22 drive arrays lessens, and the need for complex dual parity schemes is mitigated by the lower drive count.  It would ease the pressure and allow for a longer timeline on those developments, instead of being held hostage by features that will be released "sometime in the future".  There are several other vendors of storage software that will soon evolve to overcome this hurdle.  And while we all wait, again, for the slowly moving machine that is unRAID development, 3TB drives will drop below $100.  Sorry if I sound a little miffed, but it's partially from frustration over the Roadmap becoming completely useless as it has not been updated in 3 months, and the 5.x series being in Beta hell for the past month.  

[bkasten]

I see, thanks.  Interesting that the poll is almost perfectly split between 3 TB support and no 3 TB support at the moment...looks like we might not reach a clear consensus.

Seeing as how 3 of the 4 options are in favor of 3TB support in some way, I would not call it a perfect split.  Also, while there are NUMEROUS advances needed/wanted in the 5.x series, they are not as time sensitive as adding >2.1TB drive support.  AFP is only important to Mac users, AD is only important to business users.  >2.1TB support will affect most unRAID users as time rolls on.   As drive densities continue to increase, the need for 22 drive arrays lessens, and the need for complex dual parity schemes is mitigated by the lower drive count.  It would ease the pressure and allow for a longer timeline on those developments, instead of being held hostage by features that will be released "sometime in the future".  There are several other vendors of storage software that will soon evolve to overcome this hurdle.  And while we all wait, again, for the slowly moving machine that is unRAID development, 3TB drives will drop below $100.  Sorry if I sound a little miffed, but it's partially from frustration over the Roadmap becoming completely useless as it has not been updated in 3 months, and the 5.x series being in Beta hell for the past month.  

 

 

Now there is something I can agree with, mostly.

 

I would say it slightly different.

 

AFP support is important to Mac users

AD support is important to Windows users

Greater than 2.1 TB support is going to be imporant to everyone, especially to new users and the competition.

 

If competing products make it to the market with GPT support before unRAID, or unRAID take too long to support it, other people may pass this product up for other solutions. That would be a shame in my opionion. Before this post I was thinking selfishly about having support for larger HD's.

 

Now this has me thinking about it from a new user persective.

 

If I was looking NOW as a new user, at unRAID, would I pass it up because of the lack of support for "modern hard drives?"

 

Just thinking out loud...

[queeg]

I think that the "Not a priority" 4th option shouldn't say that 3TB is not part of the 5.0 release.  I'm assuming that it is.  It's not whether I never want 3TB as much as 5.0.  It's whether or not that I want to interrupt 5.0 development to fast track 3TB support.  It's not an either/or situation.  It's when and how.

 

Time is the only real factor involved.  How long before support is provided?  How much will it impact 5.0 development schedule.  It's all about time.

 

3TB is being thrust into the fray because it's price in the market has fallen enough that questions are being asked.  It's the new flashy thing.  

Scary things like "other vendors will start to support 3TB" is not real.  unRAID doesn't compete with standard RAID solutions head on.  It's different.  They cost lots more and require beefy hardware.  Who cares what they do?  It's a different market.

 

3TB may matter to everyone but what is the urgency?  Is customer good will at stake?  Will people decide to go with RAID 5 or 6 because it's not available?  I really don't think so.

 

 

EDIT:

 

What I'm saying is that I don't want to be forced into a position where I have to sell 3TB that was rushed to market and then have it blow up in my face.  If there is actually a commitment to test it then great. 

[aiden]

Scary things like "other vendors will start to support 3TB" is not real.  unRAID doesn't compete with standard RAID solutions head on.  It's different.  They cost lots more and require beefy hardware.  Who cares what they do?  It's a different market.

I have to disagree with this statement.  Other competitors include FreeNAS, FlexRAID, Drobo, etc.  All of which have been discussed freely on these forums as direct competitors.  Drobo is obviously more expensive due to its proprietary business model, but FlexRAID and FreeNAS are free.  Different market?  How so?

[queeg]

Scary things like "other vendors will start to support 3TB" is not real.  unRAID doesn't compete with standard RAID solutions head on.  It's different.  They cost lots more and require beefy hardware.  Who cares what they do?  It's a different market.

I have to disagree with this statement.  Other competitors include FreeNAS, FlexRAID, Drobo, etc.  All of which have been discussed freely on these forums as direct competitors.  Drobo is obviously more expensive due to its proprietary business model, but FlexRAID and FreeNAS are free.  Different market?  How so?

 

I'm not afraid of losing any of my customers to flexraid or freenas.

[aiden]

I'm not afraid of losing any of my customers to flexraid or freenas.

But that's for business use.  Not all unRAID users are business users.  For example, when choosing which NAS software to use, I considered all those choices before settling on unRAID.  There ARE other alternatives for unRAID that may not work for a business, but work well for home use.  That's the market Drobo, FlexRAID, and FreeNAS target, which makes them competing products.

[queeg]

I'm not afraid of losing any of my customers to flexraid or freenas.

But that's for business use.  Not all unRAID users are business users.  For example, when choosing which NAS software to use, I considered all those choices before settling on unRAID.  There ARE other alternatives for unRAID that may not work for a business, but work well for home use.  That's the market Drobo, FlexRAID, and FreeNAS target, which makes them competing products.

 

All unRAID users want reliability.  Most of my customers are buying home media servers.  What's all this talk about business use.  I'm a business, they aren't.  unRAID isn't a toy.  It's being sold for good $$money$$.  It's expected to be a quality stable product.  There are no good excuses for it not to be.

[aiden]

I'm not afraid of losing any of my customers to flexraid or freenas.

But that's for business use.  Not all unRAID users are business users.  For example, when choosing which NAS software to use, I considered all those choices before settling on unRAID.  There ARE other alternatives for unRAID that may not work for a business, but work well for home use.  That's the market Drobo, FlexRAID, and FreeNAS target, which makes them competing products.

 

All unRAID users want reliability.  Most of my customers are buying home media servers.  What's all this talk about business use.  I'm a business, they aren't.  unRAID isn't a toy.  It's being sold for good $$money$$.  It's expected to be a quality stable product.  There are no good excuses for it not to be.

When you sell unRAID to your customers as the platform of choice, you exclude all other options because of they are "toys"?  You're arguing against Drobo and the others because they aren't being sold for good money?  What does any of that have to do with supporting the next generation of high density drives?

[queeg]

I didn't call Drobo a toy.  Clearly I said unRAID isn't a toy.  The maintenance of the code base comes with responsibilities that freeware doesn't have to acknowledge.

 

Maybe you should reread my posts if you don't understand what I'm saying.

[Rajahal]

Wow, this is getting heated.

 

So far this is what makes the most sense to me:

 

AFP support is important to Mac users

AD support is important to Windows users

Greater than 2.1 TB support is going to be important to everyone, especially to new users and the competition.

 

The current limit is actually 2.2 TB, but whatever.  I agree that 3 TB support will be the biggest issue when viewing this from a new user's perspective.  If I'm a new user who wants to build a 15 TB server, 3 TB support is going to matter a lot.  If unRAID supports 3 TB drives, then I only need a 6 drive server.  If unRAID only supports 2 TB drives, then I need a 9 drive server.  That's a significant difference.  The rest of us who have already invested so much in unRAID are unlikely to dump our servers in favor of some other system that does support 3 TB drives, so we can afford to wait.  A new user likely won't feel the same.

 

Originally I was leaning towards holding off on 3 TB drive support, but now I'm seeing a strong point made for them.  It seems that it is in LimeTech's best interest (from a business perspective) to get 3 TB drive support out first and put most of the rest of the incomplete 5.0 features on the back burner.

[aiden]

Maybe you should reread my posts if you don't understand what I'm saying.

I understand what you're saying, but you're becoming increasingly hostile towards a differing point of view, and that is not my intention.  None of the items listed (AFP, AD, 3TB support, etc) have anything to do with stability or maintenance that you keep talking about.  They are feature requests, and by that standard I fail to see the relevance of a paid platform versus freeware when addressing the priorities specified in this thread.  But obviously you and I have different perspectives on the matter, and I'm sure we voted differently. 

[queeg]

Maybe you should reread my posts if you don't understand what I'm saying.

I understand what you're saying, but you're becoming increasingly hostile towards a differing point of view, and that is not my intention.  None of the items listed (AFP, AD, 3TB support, etc) have anything to do with stability or maintenance that you keep talking about.  They are feature requests, and by that standard I fail to see the relevance of a paid platform versus freeware when addressing the priorities specified in this thread.  But obviously you and I have different perspectives on the matter, and I'm sure we voted differently. 

 

Any hostility you may be feeling is purely self inflicted. 

 

The issue as I see it is whether 3TB will interrupt development of the 5.0 version.  Jumping around between code branches adds risk that defects will be introduced.  I don't know why you interjected these other freeware raid platforms but I guess it's to demonstrate that unRAID has serious competition from them.  I don't see it that way and I believe I said so. 

 

In addition, I believe that having a stable unRAID far outweighs any short term unmeasurable gain from rushing a release.  I think that everything in the 5.0 is relevant to someone, maybe not you, but someone.  Getting the AD working will actually open up a market (a paying market) that will bring in new customers.  And there are other aspects to the 5.0 version that will generate excitement that new development is ongoing.  Many people, especially new customers want nothing to do with end of life solutions.  Showing that comprehensive development is ongoing will validate the product to many new paying customers. 

 

As I've said in practically every post I made...I would benefit from and look forward to 3TB suppprt...as long as it's seriously tested and that it doesn't screw up the other development.  Trying to modify what I'm saying or interpret that differently as hostile or whatever is not going to change anything.

[squirrellydw]

I want a stable system with AFP support as as soon as possible, I don't care about 3T support at this time. I have plenty of room to expand. What is the point of a road map if it is not followed. I don't think anyone on this forum is going to switch to another platform. Also if a new user is only worried about 3t support than unraid is not for them. I looked at all of them when I started and unraid offer more flexibility than all of them.

[aiden]

Getting the AD working will actually open up a market (a [glow=yellow,2,300]paying[/glow] market) that will bring in new customers...

...Showing that comprehensive development is ongoing will validate the product to many new [glow=yellow,2,300]paying[/glow] customers...

The premise that the only paying unRAID customers are business oriented and therefore will directly benefit from AD is flawed by your perspective.  I know I have paid in full for the full version, and it's not for business use.  I'm sure I'm not the only user who has done so.

 

And to your point that comprehensive development is ongoing, I again disagree based on the fact that the Roadmap has become needlessly outdated, and the recent beta releases have been far from reliable.

 

I submit that if unRAID indeed has no competition, as you stated originally, then Lime-tech can do whatever it wants, whenever it wants, without any sense of urgency.  Somehow I doubt that is reality.

 

Trying to modify what I'm saying or interpret that differently as hostile or whatever is not going to change anything.

Goading me further by stating such things is not beneficial.  I believe this conversation is over, because for some reason this has escalated far beyond what I intended.  Again I'll leave it that you and I fail to see eye to eye on this, and that's fine.

[Rajahal]

I think we are getting a bit off track here.  There's no option in the poll for 'rush to 3 TB support at the expense of unRAID's stability'.  OK, now I'm tempted to add that as a cruel joke...but I'll refrain.  If that was an option, no one would vote for it.  Introducing instability into unRAID would ruin it for all of us - home and business users alike.

 

As has been said many times already, the issue is not if unRAID should support GPT but when.  Some strong points have been made that integrating 3 TB support into a 4.X release would be back peddling, and I'm now convinced of that as well.

 

So it seems to me that the primary point of contention at the moment is whether or not 3 TB support is a priority for the community.  We are allowed to disagree on that point, that is fine.  While a clear majority would certainly be helpful to LimeTech, it doesn't look like that is going to happen.

 

I think the point has been adequately made that 3 TB support significantly affects both business users (i.e. server builders and resellers, such as Queeg, myself, and others) and new users weighing the options between unRAID and the alternatives (both free and proprietary).  It is not as big of a priority for home users who are already running unRAID servers (especially if they are under capacity).

 

I'm sticking with my vote for option 1 because:

1) Getting 3 TB support out sooner is better for LimeTech (it is a larger factor in a new home user's decision to go with unRAID or not than are any of the other 5.0 features)

2) Anything that is better for LimeTech is better for the rest of us (because if LimeTech goes under we're all screwed )

3) I personally can wait for the rest of the 5.0 features, and I believe the majority of my customers can as well.  unRAID is currently targeted to the home user market - it is described on the LimeTech site as a solution for 'digital media storage'.  It is not (yet) targeted towards the business market, so AD is still a bell/whistle in my opinion.  AFP will benefit a lot of home users, but ultimately it is just a performance enhancement...unRAID still works in a Mac environment via SMB and NFS.  All the other stuff, well, I can wait on that too.

4) 3 TB support is better for my business.  I don't want to have to explain to my clients that they can't use those shiny new 3 TB drives that they just bought.  3 TB drives are already at a better price point than 2 TB drives (if you factor in the cost of each drive slot, etc.).  I also don't want to tell my clients that they have to buy the more expensive 2 TB drives because of software limitations in unRAID.

 

Ultimately my vote is a selfish one, as is everyone else's.

[lionelhutz]

5.0b4 was supposed to include;

AD support fixed

>2T drive support

Developer plugin supoort.

 

Now, obviously,  the roadmap is flawed and not being followed but it was Tom's intent to do all of this (AD, AFP, >3T, plugin) for the 5.0 release. So, the poll is almost fine to ask if the AD and plugin should be pushed back to a 5.1 release in order to achieve a stable 5.0 release in a shorter period of time. The 4th option in the poll is not really valid, since it would have to be more along the lines of "do not put 3T drive support into 5.0 to release 5.0 sooner" much like the 1st option is to forgo AD and plugin support in 5.0 to release 5.0 with 3T support sooner.

 

There should be no discussion about rushing a "broken" unstable release here. Limetech will have no intention of releasing such a version. Could it happen? Yes it could happen and it has happened in the past. But, the intent is to beta test and attempt to capture and fix any data damaging bugs before a version is released.

 

Peter

 

[squirrellydw]

I think you should make a new poll with two options

 

1. Put 3t in 5 now

 

2 wait till 5.1

 

This just narrows it down to what you want