unRAID Server release 4.5-beta6 available

[prostuff1]

There have been a lot of things said in this thread that i generally agree with.

 

I had no real expectations that the software would continue to be developed and that development might end at any second.  With that said I really really hope that is not the case.  I am really hoping and looking forward to version 5.0; if/when it comes out.  I purchased my key knowing full well what was currently being offered in the product and was happy with that.  The user community here is great, but without the support of JoeL., RobJ, NAS, WebooTech, BubbaQ, bjp999, and others i am not sure i would have purchased the product.  The wiki has a vast amount of knowledge in general in it and can just about answer any questions a new user might have.

 

I understand that Tom is at least checking in on the forums and has not gone MIA completely, but would it really be that hard to comment on some of the threads out there that are looking for his input (the wiki spammer related stuff comes to mind).  I know the best way to get a hold of him is via email but if you are checking the forum, just read through some of the threads that look like help may be needed.

 

An overall update on unRAID as a hole would be very welcome by, I am sure, just about everyone in the forum.  If it was to say "version 5 is being worked on and is progressing," "development has slowed for personal reasons," "development is none existent," etc., then at least we could know where things stand.

 

That's just my two cents

[WeeboTech]

@prostuiff, fair enough.

 

other comments,  The core md driver is available in source form.  if the worst case were to occur, I'm sure there is enough talent to rebuild it.

BubbaQ, Nice job on the unRAIDWeb It really started me thinking of an XMLRPC daemon for lower core level control.

 

I'm sure your opinions are noted and I think it presents a good view of what is expected for the future.

In the meantime, I think sitting tight and working together is the best approach until we hear from Tom.

He has said to me in the past, "This is the nicest most helpful community he's ever seen."

By our assistance and working together, it allows him to focus on the core.

[bubbaQ]

unRAID 5 will be a game changer.  It is also a significant ground-up rewrite of the interface, and a huge paradigm shift.

 

And don't forget... this is a fault-tolerant server.... not some bogus game or utility.  It's main purpose in life is FAULT-TOLERANCE and a bug that loses data eviscerates the whole purpose in buying fault tolerance.

 

That is a huge amount of both coding work, and designing, and "thinking" about how to do it.... as well as a lot of trepidation, and how to migrate.  It would be much easier if there was no installed base to migrate, but that installed base has to be able to upgrade.

 

Tom also has to maintain security of his IP.  Splitting commands off to a web-based UI makes it even more difficult to prevent reverse-engineering and destruction of Tom's business model. 

 

All of this work will pay off in the long run as a number of features can be split off to the web UI, and free up core development.  Most software projects like this involve a complete rewrite a few years after launch, as lessons learned take development places that were unknown of during initial development.  At some point the product starts looking like a Gypsy van with crap handing off of it, that was not part of the original design.... and going into the code to fix something elicits a "damn ... I wish I had done YYY differently in the beginning and it would be so easy to implement XXX now."

 

Once you bite the bullet to do the rewrite, it is REALLY hard (intellectually, emotionally, $$$$, etc.) to do anything on the old system.... knowing it's dead and knowing that the time spend doing anything to the old will delay the new.... and PARTICULARLY when a bug in the old is already addressed by the new.  It's like having to paint a house because the neighbors complain it is an eyesore, when you are 3 months from closing on a deal to tear it down and rebuild it new.... but you can't tell the neighbor that because if you do, then she'll sue you if you are 1-day late in closing the deal.

[Kevlar21]

bubbaQ wrote:

unRAID 5 will be a game changer.  It is also a significant ground-up rewrite of the interface, and a huge paradigm shift.

So there will be updates? You seem pretty sure that this product will materialize.

 

Once you bite the bullet to do the rewrite, it is REALLY hard (intellectually, emotionally, $$$$, etc.) to do anything on the old system.... knowing it's dead and knowing that the time spend doing anything to the old will delay the new.... and PARTICULARLY when a bug in the old is already addressed by the new.

If this is the case, then why would Tom not wish to communicate this?  To the many unanswered bug reports and posts, Tom could simply respond that yes, he is aware of this problem, and yes it will be addressed in the near feature in release x.xx. Of course there would have to be the measure of responsibility to follow through in a timely manner.

 

I've had 2 distinct experiences coming to the support forums.  I either get help on my issue, lots of advice on how to thoroughly test my problem, apply different configuration settings or tweaks and have my issue solved. Or, I post an issue, a few others post that they have similar problems, and then the thread dies because the problem hasn't been experienced by a larger portion of the community. These are the posts that Tom needs to step in and comment on, as he can't rely on the 'hero' members to have solutions to everything. I even had a performance issue recently that was forwarded to Tom graciously by one of our hero members.  The response that came back was that Tom was aware of the problem and hadn't looked into it. Tell the user that his poor performance should be 'good enough'. What? Seriously? The impression that was left with me was that the forum members that gain absolutely nothing were more than willing to help me with my issue, but the actual company that I purchased my software from couldn't be bothered to look into it.

 

BubbaQ wrote

Rule 1) Never wait for vaporware.

 

Rule 2) Don't violate rule 1.

 

If it ain't delivered and working ---- it's vaporware.

 

I have no sympathy for the wailing and gnashing of teeth in this thread (and elsewhere).  If you chose to wait, that is YOUR choice, not the developer's fault.

 

Well said, but isn't there a set of anecdotal rules that developers should follow as well?  Is it no surprise that a portion of the user base is become unsympathetic in regards to lack of support?

 

[Joe L.]

I even had a performance issue recently that was forwarded to Tom graciously by one of our hero members.  The response that came back was that Tom was aware of the problem and hadn't looked into it. Tell the user that his poor performance should be 'good enough'. What? Seriously? The impression that was left with me was that the forum members that gain absolutely nothing were more than willing to help me with my issue, but the actual company that I purchased my software from couldn't be bothered to look into it.

Please do not misquote me... The specific answer I posted was:

He thanks us for the analysis, and he did say he has noticed a difference himself, had not yet looked into it.  Other items were higher in priority, as even the slower rate of 12 MB/s was fast enough for most HD material.  He said he has not noticed any issue with his own HTPC on HD material over SMB, but realizes it might show itself on some material on some players.

 

That is very different than "not bothering to look into it"...

 

Joe L.

[bubbaQ]

Tell the user that his poor performance should be 'good enough'.

 

In programming, there are several steps... among them are:

 

Get the code working...

Get the code working and producing correct results....

Get the code working and producing correct results and running faster/better/efficiently.

 

Once some code is working and producing correct results, you move on to other areas that are NOT working or NOT producing correct results... as long as there are more important bugs elsewhere, you don't spend time making something work faster if it is working "well enough" for most purposes.  It sucks to be 4-sigma away from the norm, but that's life.

[NAS]

An awful lot of speculation here.

 

Who is right.. who knows.

 

But WHY CANT LIMETECH JUST POST AND UPDATE. I dont care how busy you can take 5 minutes to post an update.

 

I just cant get away from the fact that all this speculation is completely pointless. Hundreds of words and dozens of forum members manhours wasted to compensate for one paragraph from Limetech.

 

I mean come on lets get real.

[Kevlar21]

Joe L., it was not my intention to misquote you and if I have somehow offended I do apologize.  My intention was to point out my great experiences with forum support from excellent supporters such as yourself, and my lack of support specifically from Tom. I used my experience to illustrate the IMPRESSION that I was left with.

 

I was experiencing what I felt as poor performance results via SMB. Not a beta feature, but a core feature of the unRAID software. During the course of diagnosing the problem, the fault seemed to lie with unRAID and not my hardware.  I was given the IMPRESSION that my results were not the norm by my individual testing, the reaction of veteran members to the results in my post, as well as the general expected speeds posted in the forum and on the wiki.

 

Sure, I can admit I was disapointed that I didn't get the answer I was looking for, but I was more disapointed that there was no indication that my problem would EVER get looked at. Again, this is the IMPRESSION that is given when there is no official support presence in the forums. I can interpret Tom's message several ways.

 

1. I know about the problem, there are a few higher priorities at the moment, I've noted your case and I'll look into your problem time permitting and get back to you. I'm really busy at the moment so please be patient. If I haven't responded back in a week or so just send me a friendly reminder.  Thanks again.

 

2. I know about the problem, it's not a priority, I don't notice the problem on my system so it's REALLY not a priority, and besides your speed is 2x -3x slower than a lot of others on the forums so no worries. I would be happy with the speed you have and so should you.

 

Obviously 2 interpretations that are polar opposites, and both slanted in comparison to the original message. One clearly professional, and one that would probably piss off the user.  The truth usually lies somewhere in between. Now when the average user posts a problem in the SUPPORT forums, which LimeTech has recently not bothered to actively take part in, for many months its easy to get the IMPRESSION closer to the later interpretation.

 

When I post in the support forums, I'm always grateful to the veteran members who post suggestions and advice. I never expect that I would post problem x, and then solution y is posted hours later.  I just want Tom to know that we don't expect a perfect answer to every question we ask, just that he participate in the dialog. Sometimes a suggestion, or have you tried xxx? or read www.trythissolution.com and let me know how it works out is all it takes and the entire community benefits.

 

 

 

 

[bubbaQ]

You have apparently never been sued for a "promise" made on a website or forum post or dealt with the SEC.

 

You don't want an update... you want a commitment.  An "update" that says "There is work being done on a new version.  How much work, I can not say.  When and if it will be released is not yet known.  It may not ever be released, or it could be next week.  That's all I can say" is all I would post after the lawyers are finished.

[DrMexSS]

Hello everybody!

 

First of all, I think this is the correct place for my post because I use unraid 4.5beta6, if I've failed please let me know the correct place.

 

I've just built a Tower after some issues (bad hard disk) and I was really happy until I discovered this problem.

 

Sometimes, when I want to watch a movie it happens this:

 

1º- I power on the Tower

2º- I power the HTPC

3º- I select the movie using XBMC, which is pointing to smb://tower/PelisHD ---> User share.

4º- After 30-60 minutes of the movie, it stops. This is because the HTPC has lost the connection with Tower.

5º- I check the Tower with ANOTHER computer but the Tower doesn't respond. No http://tower, NO PING, nothing.

6º- Finally I go where the Tower is and I plug-in a keyboard (you can see in the syslog). I can access the disk, I can see in "ifconfig" that Tower has an IP, and... finally, when I type "top".... the pc that i use in 5º starts to receive the answer of the ping.

7º- Now I can return to my HTPC and continue watching the movie without interruptions.

 

I attach the syslog where is easy to find when I start to type in the Tower (Sep  8 18:45:50 Tower kernel: usb 4-2: new low speed USB device using uhci_hcd and address 2)

 

My configuration is this:

---

Unraid 4.5 beta6 - Plus key

USB: Sandisk cruzer (16gb)

 

Mobo: Asus P5Q-VM (asus)

  Intel® G45 / ICH10

  Realtek® 8111C PCI-E Gigabit LAN controllers, featuring AI Net2

 

CPU: Celeron S 430 (1x 1800 MHz)

RAM: 2 GB

HDD: 4 disks >> parity and disk1: Seagate ST31000528AS (1TB), others: 1 Hitachi 750 GB and 1 seagate 500GB

 

The router is a apple Time Capsule with DHCP, but I always give the IP: 10.0.1.2 to Tower (MAC reservation).

CAT6 cable.

---

 

Please let me know if you need more information.

 

Should I have to change the NIC? or is this a Unraid problem?

 

Thank you so much for your answers.

 

[Joe L.]

Hello everybody!

 

First of all, I think this is the correct place for my post because I use unraid 4.5beta6, if I've failed please let me know the correct place.

 

I've just built a Tower after some issues (bad hard disk) and I was really happy until I discovered this problem.

 

Sometimes, when I want to watch a movie it happens this:

 

1º- I power on the Tower

2º- I power the HTPC

3º- I select the movie using XBMC, which is pointing to smb://tower/PelisHD ---> User share.

4º- After 30-60 minutes of the movie, it stops. This is because the HTPC has lost the connection with Tower.

5º- I check the Tower with ANOTHER computer but the Tower doesn't respond. No http://tower, NO PING, nothing.

6º- Finally I go where the Tower is and I plug-in a keyboard (you can see in the syslog). I can access the disk, I can see in "ifconfig" that Tower has an IP, and... finally, when I type "top".... the pc that i use in 5º starts to receive the answer of the ping.

7º- Now I can return to my HTPC and continue watching the movie without interruptions.

 

I attach the syslog where is easy to find when I start to type in the Tower (Sep  8 18:45:50 Tower kernel: usb 4-2: new low speed USB device using uhci_hcd and address 2)

 

My configuration is this:

---

Unraid 4.5 beta6 - Plus key

USB: Sandisk cruzer (16gb)

 

Mobo: Asus P5Q-VM (asus)

  Intel® G45 / ICH10

  Realtek® 8111C PCI-E Gigabit LAN controllers, featuring AI Net2

 

CPU: Celeron S 430 (1x 1800 MHz)

RAM: 2 GB

HDD: 4 disks >> parity and disk1: Seagate ST31000528AS (1TB), others: 1 Hitachi 750 GB and 1 seagate 500GB

 

The router is a apple Time Capsule with DHCP, but I always give the IP: 10.0.1.2 to Tower (MAC reservation).

CAT6 cable.

---

 

Please let me know if you need more information.

 

Should I have to change the NIC? or is this a Unraid problem?

 

Thank you so much for your answers.

 

See this answer for some things to try (basically, they moved the "IRQ" the network card was assigned to one less used) : http://graag.blogspot.com/2007/12/netdev-watchdog-eth0-transmit-timed-out.html  It is a apparently a bug with the driver for some realtek chipsets.  Worst case, install a different network card.

 

 

[WizADSL]

You have apparently never been sued for a "promise" made on a website or forum post or dealt with the SEC.

 

You don't want an update... you want a commitment.  An "update" that says "There is work being done on a new version.  How much work, I can not say.  When and if it will be released is not yet known.  It may not ever be released, or it could be next week.  That's all I can say" is all I would post after the lawyers are finished.

 

All we (I) want is a response from LIMETECH, the content of that response is up to them as long as it pertains to the status of the unraid project.  Plenty of other projects that are in much wider use and under more public scrutiny such as MySQL, Apache, Firefox and others make statements about current and future development apparently without fear of legal action.  Limetech could put one of those "forward looking statement" disclaimers on the post, I don 't care.  I just want to know the status of development, perhaps that is what most of us want.  If your post today at 12:14pm had come from Limetech for example, I would be a happy camper.

[cwr]

i sent through some queries about a few bugs to tom last night, he said he'll be back on the forums soon

[SSD]

How about a contest to see how long "soon" is (i.e., how long it will take Tom to post in the forum again).

 

I'm guessing 2010.   

[smino]

Ouch you guys! Give Tom a break. Summer is not over yet!

[Joe L.]

Ouch you guys! Give Tom a break. Summer is not over yet!

True, and he has taken breaks before

 

 

In correspondence to me he has also said he would be back on the forums soon. 

 

Joe L.

[Kevlar21]

I'm glad to hear that Tom will be back to a more active role on the boards.  A lot of unRAID users are passionate about this product, and with passion comes emotion. There has been a lot of skepticism as well as a lot of optimism on the future of unRAID as well as as the existing hiccups of its current revisions. Given the absence of any official clarity from Tom as of late leads to a lot of speculation. If I am to interpret the graph posted by Joe L., it would appear that LimeTech participation on the forums in somewhat intermittent at times. As it seems that unRAID development will continue and evolve my question to the forums is simply this.

 

Given unRAID's success and growing userbase, what is the expected or acceptable level of official support from LimeTech/Tom?

 

[NAS]

Ouch you guys! Give Tom a break. Summer is not over yet!

True, and he has taken breaks before

 

In correspondence to me he has also said he would be back on the forums soon. 

 

 

Abolutely TOm should be free to come and go as he pleases. Limetech LLC however should ALWAYS be here. That is the burden of starting a company, making money and the duty of care for customers.

[bubbaQ]

NAS, how many copies of unRAID have been sold?  Without knowing that, there is no way ANYONE can make such demanding statements about what "duty" there is. 

 

Let's say there are 10,000 sales.... and how many are bitching and moaning in the forum?  Less than 0.5%.  Frankly, if it was me, I'd gladly flush 0.5% of complaining purchasers to be rid of 95% of all my headaches and support problems.

 

Let's say there are only 500 sales.  At that rate, the whole thing isn't worth my time, much less putting up with the level of complaints from 10% of purchasers... I'd pack it in.  It CERTAINLY does not support paying even one part-time employee.

 

Some of the "bugs" complained about such as write speed, are not going to be "fixed" because they CAN'T be.... they aren't bugs they are system limitations.  It is not a "bug" that  that a Honda Accord can't go 400 mph.... it is a consequence of the laws of physics, the available materials, and the design constraints of the vehicle.  If you want write speed, go to a striped RAID5.... or use a PCIe SSD for your parity drive.  1TB will set you back a couple of grand, but you will have some SMOKIN' write speed (up to the raw write speed of your data drive).  There, problem solved.

 

Tom obviously LIKES coding unRAID.... cause it sure isn't worth the money.  WE are alll the beneficiaries of Tom's willingness to get squat in return for his work... I've known (and been) many coders that get to do something they like doing.  If you take away the "likes" part due to creating overhead, a developer will quickly lose interest in the project.  Money is not a motivator for many good coders as almost all of us could easily make more money somewhere else -- most don't jump ship for the simple reason of liking the place we are at (figuratively) rather than being a monkey in a cage getting more money and hating the drudgery.

 

unRAID is a product that is worth many times what it sells for to some people... and not worth 1/10 of the price to others.  I think most complainers should simply find a different product.  Perhaps if they tried that, they'd discover what a deal unRIAD is. How many "alternatives" get mentioned in the forums, and are shot down as inferior or too expensive?  But don't forget the unattainable triad (good, fast, cheap)... pick 2... not only from an alternative, but also from unRAID.

[Kewjoe]

I agree 100% with BubbaQ and others along the same line. Tom/Limetech has been off the forums for so long, yet is still responding to emails in a reasonably time frame. We should stop treating these forums like some official 24/7 support forum and realize they are here for our added benefit. Tom could easily just close the forum down and just support his product via email. Many products and companies have forums that don't have employees responding at all or in any constant basis.

 

Like someone said earlier in the thread, when you paid for this product, you paid for how the product performed that day, not how it'll evolve over the next few years. I'm glad Tom has thick skin, because the way some of you act, I don't know if I could put up with it. I don't think that Tom became an instant millionaire making this product. At some point the bickering will hit a nerve and the willingness to innovate will disappear. I hope that doesn't happen.

[NAS]

NAS, how many copies of unRAID have been sold?  Without knowing that, there is no way ANYONE can make such demanding statements about what "duty" there is.  om an alternative, but also from unRAID.

 

I dont care how many have been sold. The forum is being presented as commercial support which it shouldnt be as it isnt. Dont get me wrong I still love unRAID and the community but the forum is a comminity forum not as it is presented... a commercial support forum.

 

...We should stop treating these forums like some official 24/7 support forum and realize they are here for our added benefit....

 

Then the obvious solution is for Limetech LLC to stop listing the forum in the official sales site as a means of commercial support. Sorry I am not trying to be picky but you cant have it both ways, profit from sales and users doing commercial support. Or put a better way loads of people are prepared to help out on the forums, and that is the way it should be, but it should not be with thec omplete exclusion of the company whos product is being supported.

 

The big 10 users could quite easily go away and where would that leave users? OK so they are not going to go away but I refuse point blank to believe that Limetech LLC cannot spare 1 man hour in six months  to assist in community support.

[teamhood]

Wow - I didn't know that so much has been happening with unRAID the past couple of months. I was one of the early adopters of unRAID back when it was first announced at AVSForums and I recall when Tom went MIA for quite a few months. I've been very happy with my array and everything has been working great for over a year now and  I have not needed to do any upgrading.

 

I'm sure Tom will be back soon... thankfully there are some amazing users here who are willing to help, otherwise I believe this product would fold....

[Kewjoe]

Then the obvious solution is for Limetech LLC to stop listing the forum in the official sales site as a means of commercial support. Sorry I am not trying to be picky but you cant have it both ways, profit from sales and users doing commercial support. Or put a better way loads of people are prepared to help out on the forums, and that is the way it should be, but it should not be with thec omplete exclusion of the company whos product is being supported.

 

The big 10 users could quite easily go away and where would that leave users? OK so they are not going to go away but I refuse point blank to believe that Limetech LLC cannot spare 1 man hour in six months  to assist in community support.

 

 

To me, the wording is very clear. "Many regular posters eager to help out, including several long-time users of unRAID OS" as a paying customer, that set my expectations. Also noted is, you can contact Lime Technology directly with any technical support issues or questions. It is very clear to me how support is structured.

 

The top 10 users, I gather stay and help because they'd rather Tom develop unRAID 5.0, rather than spend time answering/fixing issues with 4.4/4.5 which will be obsolete soon. But that is purely my assumption and I don't speak for them.

 

If this bothers you so much, if you feel like you bought the product under the understanding that Tom/Limetech would be monitoring these forums on a daily basis, I don't know what to tell you. Perhaps it best to ask Tom for a refund.

[WizADSL]

NAS, how many copies of unRAID have been sold?  Without knowing that, there is no way ANYONE can make such demanding statements about what "duty" there is. 

 

Let's say there are 10,000 sales.... and how many are bitching and moaning in the forum?  Less than 0.5%.  Frankly, if it was me, I'd gladly flush 0.5% of complaining purchasers to be rid of 95% of all my headaches and support problems.

 

Let's say there are only 500 sales.  At that rate, the whole thing isn't worth my time, much less putting up with the level of complaints from 10% of purchasers... I'd pack it in.  It CERTAINLY does not support paying even one part-time employee.

 

Some of the "bugs" complained about such as write speed, are not going to be "fixed" because they CAN'T be.... they aren't bugs they are system limitations.  It is not a "bug" that  that a Honda Accord can't go 400 mph.... it is a consequence of the laws of physics, the available materials, and the design constraints of the vehicle.  If you want write speed, go to a striped RAID5.... or use a PCIe SSD for your parity drive.  1TB will set you back a couple of grand, but you will have some SMOKIN' write speed (up to the raw write speed of your data drive).  There, problem solved.

 

Tom obviously LIKES coding unRAID.... cause it sure isn't worth the money.  WE are alll the beneficiaries of Tom's willingness to get squat in return for his work... I've known (and been) many coders that get to do something they like doing.  If you take away the "likes" part due to creating overhead, a developer will quickly lose interest in the project.  Money is not a motivator for many good coders as almost all of us could easily make more money somewhere else -- most don't jump ship for the simple reason of liking the place we are at (figuratively) rather than being a monkey in a cage getting more money and hating the drudgery.

 

unRAID is a product that is worth many times what it sells for to some people... and not worth 1/10 of the price to others.  I think most complainers should simply find a different product.  Perhaps if they tried that, they'd discover what a deal unRIAD is. How many "alternatives" get mentioned in the forums, and are shot down as inferior or too expensive?  But don't forget the unattainable triad (good, fast, cheap)... pick 2... not only from an alternative, but also from unRAID.

 

I would not be surprised if the number of sales is under 500, but the point here is not the number of sales, it is that we are talking about SALES and not DOWNLOADS.  The moment you decide the sell SOFTWARE, that means you are getting paid for your work and it is not merely a hobby (or at least you shouldn't treat it as one).  I develop software of a living as well, and if I take money from someone to write software for them, I accept a certain obligation to that project.  I should also say that one of the most rewarding aspects of developing for ME is to see my software used, if I were Tom, I would be incredibly proud of my achievement and that so many people use it on a daily basis.  If you want to develop software as a hobby and don't want to have any support obligations then just give it away (as so many people do) or clearly state that it will not be supported.  I would agree completely that unraid is worth MANY times what is has earned, the contributions YOU alone have made are probably worth more on their own and there is no other product quite like it, THAT IS WHY WE WANT TOM BACK, WE DON'T WANT AN ALTERNATIVE, WE WANT UNRAID AND NOT SOMETHING ELSE.  I looked at all the other options BEFORE I bought unraid.

 

Where you talk about "bitching and moaning" users, I ASSUME (I apologize in advance if I am wrong) that you are talking about users like me that are questioning the project's status and also why Tom is not on the forums.  I don't see how either of these questions (asked in a civilized manner) could put one in the "bitching and moaning" category.

 

On the separate note of write speed, I can't speak for others, but my array is comprised entirely of 1.5GB Seagate barracudas, even at the innermost (slowest) tracks, the write speed is 77MB/s, the cpu in my box is a dual-core (Intel Core 2) 2.8ghz.  My unraid write speed is about 12MB/s avg (in Honda Accord terms that's about 25mph) regardless of whether it is to a user share or a disk share.  I have no idea if the issue is with any of the unraid core or the md driver or the kernel.  But I can be sure it is NOT a physics limitation but rather it is likely to be a software issue SOMEWHERE.

 

Please don't take this as an attack on Tom or Limetech or anyone else, as that is not what it is.  This is just MY personal reaction to the statements made on this issue.  I hope I didn't break any eggshells.

[bubbaQ]

Addressing your speed issue....

 

Reading and rewriting the SAME sector on rotating magnetic media is a performance disaster.... that's the laws of physics.

 

Back in the day write speed could not keep up with rotational speed, so we had interleaving, and an improper interleave setting would send your disk performance in the toilet.  Having to read and rewrite the same sector is similar to having a too-aggressive interleaving setting -- you spend a LOT of time waiting for the disk to spin back around (as opposed to waiting for the head to seek).  This is why high rotational speed drives as parity affect writes as much as they do.

 

The only way to improve this is some sort of aggregator/elevator method of reading and rewriting, but that itself is fraught with peril in this type of environment.... not to mention would be a complete ground-up rewrite of the core.

 

If you want proof that the bottleneck is the disk and not unRAID, replace parity with a PCIe SSD.  I would almost be willing to rent one just to prove it to the doubters.