noone has passed the limits of Gigabit ethernet performance (125MB/s) with UNRAID

 

Uhhh... the limit on CSMA/CD transport like gigabit Ethernet is only 1/4 to 1/3 of the theoretical limit, or about 40MB/s.

 

 

Really?  So why do these NAS systems demonstrate average read/write performance at around 85MB/sec on gigabit ethernet?  These use IOZONE benchmarks and average various transfer sizes that I had previously posted for UNRAID on my old system:

 

http://www.smallnetbuilder.com/component/option,com_nas/Itemid,190

I think the Supermicro ITX board posted earlier is going to be quite the lil performer for an environment like unRAID.  Thanks to Aiden for posting that link.

No problem, and thanks for posting a response that was more helpful then what I was about to post.  

I believe that no one has done the benchmarking tests you are looking for because we all know (based on how unRAID works) that the CPU and RAM make little to no difference.  As previously stated, the bottlenecks will always be network speed, bus speed, and drive speed.  There have been plenty of benchmarks performed on those parameters, since those are the ones that matter.

 

If you want to perform your own benchmarks comparing a dual core atom and an i7 to appease yourself, go for it.  However, I'm confident that we can all predict the answer for you - it will not matter, so don't waste an i7 on unRAID.  Some people here are still using Pentium III's, for godsakes!

 

The crux of this is that file transfers do not tax a CPU at all.  Just as a faster CPU will not affect your transfer speeds to an external hard drive, the same applies here.  Think of unRAID as a massive, networked, 'smart' external hard drive.  It does require a CPU to run, but it rarely is used (I believe it is only used during writes and parity builds/checks, but I may be wrong there).

Perceived boost in speed.

 

The initial transfer is actually faster to a temporary unprotected holding place.

Then later, the data is moved to the protected array at the normal. data to protected array,  transfer rate.

I knew that after reading the wiki, just testing.    Thanks for the clarification.

So why do these NAS systems demonstrate average read/write performance at around 85MB/sec on gigabit ethernet?

 

Not on Windows on a typical home setup with class 2 switches and 2ms latency.

 

What kind of data are you looking for?

 

 

I was hoping to see something like the IOZONE benchmarks I referenced earlier which eliminate operating system optimizations and also averages several transfer sizes.  Seems to be a good way to document "average" read and write performance if you read the test procedure used on the "SmallNetBuilder" website below.  He also has a great database to compare the results of UNRAID to some of the more expensive commercial solutions which are available.

 

http://www.smallnetbuilder.com/content/view/30682/77/

Not on Windows on a typical home setup with class 2 switches and 2ms latency.

 

I do not have a typical home network with a class 2 switch and 2ms latency.  There are many gigabit switches available for < $200 which have far better performance than this.

 

Also, for others:

 

When benchmarking, I normally connect the UNRAID server directly to my desktop PC to measure the potential of the NAS.  This takes any network traffic or switch limitations out of your measurements to better compare to others.

That's like testing a car on a racetrack with no traffic, when you intend to USE the car in rush-hour traffic.

 

I can configure a system that will work light lightning in a controlled environment, but I can't USE the damn thing in that environment.

 

So what type of switch do you have, and what latency do you have across each collision domain?  (I really want to hear this!)

The guy just wants someone to do some benchmarks for him, and post them in a (SmallNetBuilder?) format, whatever that means.

 

 

The guy just wants someone to do some benchmarks for him, and post them in a (SmallNetBuilder?) format, whatever that means.

 

 

 

If you have a better suggestion to compare performance, please advise.  If you have nothing of value to contribute, please move along.

 

If you have a better suggestion to compare performance, please advise.  If you have nothing of value to contribute, please move along.

 

 

Interesting.

 

Just what value have you contributed?

 

 

 

If you have a better suggestion to compare performance, please advise.  If you have nothing of value to contribute, please move along.

 

 

Interesting.

 

Just what value have you contributed?

 

 

 

I've provided you and bubbaQ with something to bitch about.  Isn't that something? 

 

Please go away now so any others who share my interest can have a productive discussion.

unRAID is a proprietary operating system - it is not open source.  

I disagree.  unRAID is not an operating system. 

 

unRAID is a proprietary network storage application (the management web-interface and user-shares file system) built on a open source operating system. It is a stripped down Slackware 12.2 Linux. 

 

The source code for the unRAID driver based originally on the "md" device in Linux and is GPL, and included in the lime-technology distribution.  The management utility and user-shares are not open source and not based on GPL source.

 

If you have a better suggestion to compare performance, please advise.  If you have nothing of value to contribute, please move along.

 

 

Interesting.

 

Just what value have you contributed?

 

 

 

Wow, this is turning into a flame war very quickly.  The dramallama has no place here.

 

GaryMaster: Please just listen to what the experienced unRAID users in this thread have already said, especially the Hero members.  They know what they are talking about.  For the last time, the benchmarks you are asking for are useless and don't apply to the unRAID environment.  If you want speed over data protection and ease-of-use, then unRAID is not for you.

 

Everyone else: Keep in mind that GaryMaster is clearly new to unRAID and comes from the world of performance-at-all-costs.  He is used to the CPU and RAM being critical in any situation.  It took me a while to get used to this concept when I first discovered unRAID as well (though, admittedly, I didn't get belligerent over the idea!).  If he had a front wheel drive sports car, he would probably want to perform benchmarks on various trunk-mounted spoilers. (sorry, I just had to work that in there somewhere )

 

I vote this thread be closed/locked.  There is nothing productive that is going to come of this.

unRAID is a proprietary operating system - it is not open source.  

I disagree.  unRAID is not an operating system. 

 

unRAID is a proprietary network storage application (the management web-interface and user-shares file system) built on a open source operating system. It is a stripped down Slackware 12.2 Linux. 

 

The source code for the unRAID driver based originally on the "md" device in Linux and is GPL, and included in the lime-technology distribution.  The management utility and user-shares are not open source and not based on GPL source.

 

Fair enough, I'll take your word for it.

I vote this thread be closed/locked.  There is nothing productive that is going to come of this.

 

Ditto.  No more feeding the trolls.

I vote this thread be closed/locked.  There is nothing productive that is going to come of this.

Unfortunately I have to agree.  Gary needs to do his own benchmarks to believe what the experienced users are telling him.  He's a numbers guy, and I honestly think that even if we did give him numbers they would be considered suspect.  There's no harm in trying it Gary.  There's no opposition to your performing extensive tests on multiple platforms, just recognize that most of us haven't done so because it's just not that impactful in the end.  The numbers you find on here are almost always parity rebuild times, drive speed, throughput, and unRAID version announcements.  

 

Everyone else: Keep in mind that GaryMaster is clearly new to unRAID and comes from the world of performance-at-all-costs.  He is used to the CPU and RAM being critical in any situation.  It took me a while to get used to this concept when I first discovered unRAID as well (though, admittedly, I didn't get belligerent over the idea!).  If he had a front wheel drive sports car, he would probably want to perform benchmarks on various trunk-mounted spoilers. (sorry, I just had to work that in there somewhere )

 

I vote this thread be closed/locked.  There is nothing productive that is going to come of this.

 

I have actually been using UNRAID for 4 years through many versions of the software.  All I have been asking for are some quantified results showing the CPU/Memory/Chipset having no measureable impact on gigabit transfer performance.  For unknown reasons some seem to take personal insult from this question.  Citing a specific comparative measurement to illustrate that all processors deliver the same performance would satisfy me and I would be happy to move on to a discussion of systems that do matter (hard drives, controllers, bus, etc).  All I am trying to do is advance the best that can be done with UNRAID for a reasonable amount of money.  It seems to me that this is something good that benefits the entire community.

For unknown reasons some seem to take personal insult from this question.

 

The reason is not "unknown", dude.  Just look at your attitude:

I've provided you and bubbaQ with something to bitch about.  Isn't that something?

 

So far numerous people have given you plenty of useful information to think about. 

You just seem to not see this.  And you keep rejecting it.

 

Therefore there is no use continuing this thread.  Might as well close it.

 

 

I have actually been using UNRAID for 4 years through many versions of the software.  All I have been asking for are some quantified results showing the CPU/Memory/Chipset having no measureable impact on gigabit transfer performance.

We have told you it does not.  There are many people here that have run many different setups and they have told you that it makes no difference what so ever.  I can also attest to this as I have been running an Abit AB9 Pro with the processor overclocked, underclocked, and at stock voltage.  It made NO difference what so ever in day to day use.  Any modern CPU will be more then enough (including most Atom processors) for the demands of unRAID. 

 

For unknown reasons some seem to take personal insult from this question.  Citing a specific comparative measurement to illustrate that all processors deliver the same performance would satisfy me and I would be happy to move on to a discussion of systems that do matter (hard drives, controllers, bus, etc).

There is the User Benchmarks Page that we have tried to flesh out.  It mostly documents the speed and time it takes to do a parity check and how many drives were involved.

 

All I am trying to do is advance the best that can be done with UNRAID for a reasonable amount of money.  It seems to me that this is something good that benefits the entire community.

I'm not saying unRAID cant get faster but we are at the mercy of rotational speed and bus bandwidth... there is not a whole lot that can be done about that... short of letting new hardware come out.  Fasters disks will speed up the process and wider busses will help, but unRAID does not gain anything from a faster processor.

 

We have told you it does not.  There are many people here that have run many different setups and they have told you that it makes no difference what so ever.  I can also attest to this as I have been running an Abit AB9 Pro with the processor overclocked, underclocked, and at stock voltage.  It made NO difference what so ever in day to day use.  Any modern CPU will be more then enough (including most Atom processors) for the demands of unRAID.  

 

There is the User Benchmarks Page that we have tried to flesh out.  It mostly documents the speed and time it takes to do a parity check and how many drives were involved.

 

I'm not saying unRAID cant get faster but we are at the mercy of rotational speed and bus bandwidth... there is not a whole lot that can be done about that... short of letting new hardware come out.  Fasters disks will speed up the process and wider busses will help, but unRAID does not gain anything from a faster processor.

 

 

Thanks, Man.  I did spend some time on that benchmark page you sent.  I believe those tests are all local performance to the server and not network transfer performance related.  I had also read several posts here explaining differing people used different methods on those submissions and that the results were not really valid/comparable.  I'm sorry that I'm so anal about this stuff, but I'm a veteran Electrical Engineer and it's in my blood!

 

For any who are interested I will wait for the i7 processor I have on order to arrive and then give it a shot against the dual core Atom board with a pair of WD Raptor 300 drives installed (one parity and one data - I believe this is still one of the fastest rotating drives, so it will show us the best practical speed we could expect).  The i7 will have DDR3 tripple channel memory and the atom will have DDR2.  Everything else will be the same between the two systems.

 

I will test transfer performance directly coupled to my desktop PC, then through my gigabit switch and then again through a more typical $50 8 port gigabit switch with no other network traffic present.  Based on all of the feedback here, I should see negligable differences between the systems (the concensus seems clear on this point), but I will still be interested to find out.  

 

If anyone is interested in the results, please send me a message and I will shoot them back to you directly (I will no longer be monitoring this thread since it seems to be a trainwreck!).

So far, I haven't learned much

 

That, is an understatement.

 

Adding CPU horsepower past a Pentium III is WASTED effort.  It has no appreciable impact on unRAID read or write performance.  If you can't understand that NATIVELY and without having to "prove" it with bench testing, then you have, indeed, not learned much.

 

 

Actually unRAID is CPU bound. Simply because it uses user space shfs under the hood.

I can feel it on my old single core CPU (AMD Athlon 3200+) - load average goes high during copying process.

 

A high load average does not mean that it's CPU bound. It's mostly waiting on I/O to complete.

I'm not usually one to complain. I understand the above saying about beating a dead horse is a old one, but I could have done without the graphic. As a horse owner, I find it disturbing. In fact one of my horses (my personal horse is a grey) is pure white.

 

Again sorry to complain but imho it is in bad taste..

 

Phil

For any who are interested I will wait for the i7 processor I have on order to arrive and then give it a shot against the dual core Atom board with a pair of WD Raptor 300 drives installed (one parity and one data - I believe this is still one of the fastest rotating drives, so it will show us the best practical speed we could expect).  The i7 will have DDR3 tripple channel memory and the atom will have DDR2.  Everything else will be the same between the two systems.

 

Actually I think a more practical test is with Seagate 7200 RPM 1.5TB drives.

From what I've read, these have been benchmarked as fast, if not faster, then raptors.

In addition, I doubt people will ever put a raptor in an unraid system, but many will use the larger drives (5400,5900 & 7200 rpm).

 

Trainwreck? Hmm.. I did not think so. There was just a bit of tudes floating around.

The thread does not need to be locked. Part of the discussion had some points.

I would like to see Gary back up the theories with numbers from the familiar benchmarks when hardware is present. My own anecdotal tests revealed that a 1.6 and a 2.6 CPU provided the same throughput.

An Atom vs an i7 will be an interesting test.

 

I remember Tom saying that this was the nicest community he's had the pleasure of working with.

So for those who post/respond. Please re-read your replies and consider others before clicking that post button.

If there's any chance that someone may take offense to what or how something is said. you can be sure of a similar response back.

 

Someone once posted in another forum.  "Attack idea's, not people".  Consider that in the discussion.