Think I'm leaving Unraid

[dschur]

While my unraid has been a reliable, incredible workhorse, I haven't put a lot (or really anything) into it over the years and it's a bit of a monster of wires and power supplies these days.

 

To put it in perspective, the largest drive in it now is 750GB, it has 2 750GB, 3 500GB, 1 320GB, 4 300GB, and 3 250GB drives in it. 13 drives total, and 2 250watt power supplies Most of the drives are PATA (remember those? ). Thing probably costs me $400/year in electricity ;-)

 

Today my unraid lost a drive (first in about 4 years so that ain't bad), it'll run unprotected, but I'm faced with buying a drive to keep my heap running or migrating. I've shut it down until I can start moving stuff off. I owe unraid a great debt for dying gracefully and giving me the time to get my data off with a dead drive.

 

Also, over the years I have migrated to all Macs as my client devices (2 Macbook Pros, Macbook Air, and 3 minis with Plex for media servers/workstations).

 

The AFP support, time machine, native Mac integration, simplicity, and overall cleanliness of Drobo has lured me over. Sorry Tom I got tired of waiting for AFP. I placed an order for a drobo FS an 2TB drives

 

If anyone  needs a CM stacker case, unraid pro keys, mobo/cpu/ram/etc (forgot the model but it was the intel one tom insisted on in the start) shoot me a note, cause I'm out. I'm in Chicago if anyone wants the box, not really worth shipping as I think it weighs about 70 pounds.

 

 

 

[bkasten]

While my unraid has been a reliable, incredible workhorse, I haven't put a lot (or really anything) into it over the years and it's a bit of a monster of wires and power supplies these days.

 

To put it in perspective, the largest drive in it now is 750GB, it has 2 750GB, 3 500GB, 1 320GB, 4 300GB, and 3 250GB drives in it. 13 drives total, and 2 250watt power supplies Most of the drives are PATA (remember those? ). Thing probably costs me $400/year in electricity ;-)

 

Today my unraid lost a drive (first in about 4 years so that ain't bad), it'll run unprotected, but I'm faced with buying a drive to keep my heap running or migrating. I've shut it down until I can start moving stuff off. I owe unraid a great debt for dying gracefully and giving me the time to get my data off with a dead drive.

 

Also, over the years I have migrated to all Macs as my client devices (2 Macbook Pros, Macbook Air, and 3 minis with Plex for media servers/workstations).

 

The AFP support, time machine, native Mac integration, simplicity, and overall cleanliness of Drobo has lured me over. Sorry Tom I got tired of waiting for AFP. I placed an order for a drobo FS an 2TB drives

 

If anyone  needs a CM stacker case, unraid pro keys, mobo/cpu/ram/etc (forgot the model but it was the intel one tom insisted on in the start) shoot me a note, cause I'm out. I'm in Chicago if anyone wants the box, not really worth shipping as I think it weighs about 70 pounds.

 

 

 

 

 

I find your post interesting, as a new user to unRAID. I have some questions for you, if you are willing to answer them.

 

I too am a Mac user, and many of the systems I have in my house are Macs. At work I am a systems admin, and have a certain expertise with file serving for Macs, at least I was paid for it.

 

As such I have tested/ran servers for AFP on a lot of platforms, both at work and at home. Because I have this experience, I have some definite opinions about serving AFP, and what the costs are and the advantages/disadvantages.

 

What are you looking to get by having a server capable of serving AFP? (Like the Drobo)

 

Bruce

[dschur]

Hi  bkasten, I'm more than happy to give any insights I can.

 

First and foremost, you seem to have inferred that my primary reason from migrating from unraid to drobo is AFP support. Sorry if I left that impression, as the protocol really did not factor into my decision much, if at all.

 

I should assure that lack of AFP support  is absolutely not the primary factor in moving on. My primary reason is I have a bunch of old junk from a hardware perspective and while unraid has allowed me to combine it and keep it running together, I am starting to see HW failure and very much worried it will be a cascading effect. I can either reinvest in it or move to something else. I have not touched this stuff in 4-5 years which is amazing in the technology lifecycle.

 

As for AFP, I run Leopard Server in my small home environment. I hope/suspect is a certain elegance and simplicity to it throughout as opposed to SMB (i've never had the ability to check). I generally view Samba as equivalent to AFP, and I am looking forward to see if having an AFP server will offer me any advantages.

 

That said, AFP support has been an unfulfilled promise for me on unraid to date. I have heard for  few years now that it is around the corner on unraid, I have patiently waited but have nothing to show for it, so for me, given my (personal) hardware issues, it's time to move on. It's really a choice of where I invest my hardware dollars combined with the fact that I have a bunch of now nearly 5 year old unraid hardware.

 

The direct corollary is having a 15 year old used car. repair may be a good idea, but sometimes its best to trade it in and move on. Having a complete set of the latest and greatest hardware will probably serve me well for the future as well.

 

 

 

[madpoet]

See... after 4 years, I tore my orginal down and just rebuilt it in the CM Stacker case   Still cheaper than any of the alternatives.  But I do not have a Mac.

[Joe L.]

See... after 4 years, I tore my orginal down and just rebuilt it in the CM Stacker case   Still cheaper than any of the alternatives.  But I do not have a Mac.

After 5 years I built a new server using the C2SEE motherboard and all 7200 RPM SATA drives.  It has fewer disks, and nearly twice the capacity.

 

I've purposely kept the parity disk on the older server at 1TB, since parity calculations would take twice as long if I upped it to a 2TB drive, and they already take a long time.

Most of the disks are PATA, with a two 750Gig drives and the remainder 500Gig and smaller.

 

I cannot see paying a premium for a PATA drive if one of the IDE drives were to fail.  I would need to use an SATA drive.  I do have a 4-port SATA controller card but like all the other disks, it is on the PCI bus.    I'll have to replace the IDE Icy-Dock enclosures when that time comes.  As others have stated the CM Stacker case is VERY nice.

 

Soon I will make my newer server the one for serving most of the media in my house and begin the process of installing a newer motherboard into the old CM case... and have been giving the transition some thought.  I found a great deal on a dual CPU Xeon based motherboard.  It is PCI-X and it can run the PCI bus at 66 MHz, so it will be twice the speed even with the older IDE drives.  I just need to find the time to make the transition.

[SSD]

I like the name of your thread "Think I'm leaving Unraid".  Sounds like you're second guessing your decision already. 

 

I too started out with mostly an IDE based array (250G/300G), but most expansion was SATA.  I haves since thrown out all the IDE relics now, and just recently removed last of the 750G, leaving just 1T, 1.5T and 2T drives. 

 

I would recommend holding on to that unRAID key.  Once that Drobo melts all your drives, you'll be dying to come back. 

[bkasten]

Hi  bkasten, I'm more than happy to give any insights I can.

 

First and foremost, you seem to have inferred that my primary reason from migrating from unraid to drobo is AFP support. Sorry if I left that impression, as the protocol really did not factor into my decision much, if at all.

 

I should assure that lack of AFP support  is absolutely not the primary factor in moving on. My primary reason is I have a bunch of old junk from a hardware perspective and while unraid has allowed me to combine it and keep it running together, I am starting to see HW failure and very much worried it will be a cascading effect. I can either reinvest in it or move to something else. I have not touched this stuff in 4-5 years which is amazing in the technology lifecycle.

 

As for AFP, I run Leopard Server in my small home environment. I hope/suspect is a certain elegance and simplicity to it throughout as opposed to SMB (i've never had the ability to check). I generally view Samba as equivalent to AFP, and I am looking forward to see if having an AFP server will offer me any advantages.

 

That said, AFP support has been an unfulfilled promise for me on unraid to date. I have heard for  few years now that it is around the corner on unraid, I have patiently waited but have nothing to show for it, so for me, given my (personal) hardware issues, it's time to move on. It's really a choice of where I invest my hardware dollars combined with the fact that I have a bunch of now nearly 5 year old unraid hardware.

 

The direct corollary is having a 15 year old used car. repair may be a good idea, but sometimes its best to trade it in and move on. Having a complete set of the latest and greatest hardware will probably serve me well for the future as well.

 

 

 

 

Thank you for your reply. I hope I understand your views better now.

 

I completely agree with retiring old HW. At some point, fixing it can be downright dangerous, data-wise. In the places I work, among the SA's we smirk and say "Hardware will eventually fail, Software will eventually work". It's a way of poking fun at the programmers we often work with. 

 

Getting back to the reason I was asking was I have just come to the product after researching the competition rather vigorously. I have looked over a number of NAS devices, including DROBO, and after reading the reviews, I was kind of disappointed.

 

I have run Apple Servers, the nice rack servers, and boy was I a happy admin. Apple on Apple is really a great, fast product for speed, reliability, expandability, etc. I have never seen anything better (for serving the Mac users). I have also run non-Apple AFP products, both at work and at home and frankly been seriously put off. Basically, if it's not Apples own AFP, I have not had much success with it. Naturally, this wonderful server set up was at a place that I worked, and was a bit pricey for a home setup.

 

Part of my coming to this product is I wanted to have what you have had. I want this thing to run for 4-5 years with minimal input from me. I want to avoid the quirky behavior, the data loss, and the outright failures that I have had in the past. I wanted to avoid the proprietary hardware and the very limited expansion that most of the other products offer. I came to unRAID in hopes that when I reach the point where you are now, I could replace what has become junky old HW with more shiny new HW, and keep going with minimal investment. I was hoping that I could get long term service out of this.

 

I keep reading over your post. I cannot find a complaint about the product, other than the AFP part. I guess that is why I latched on to that as the issue. As much as I like AFP as a protocol, it is (mostly, in my opinion) proprietary. *sigh*

 

I guess I was struck by the irony that you have achieved what I want to, but you are dissatisfied.

 

Peace

 

Bruce

[unraided]

Wow, that's heavy. You're leaving the pact over the The Apple Filing Protocol? Me thinks that we might have a 'The Prodigal Son' moment in the future to come . One way to ditch your Mac woes, get a Wintel box, . As mentioned before in this post, I'd definitely keep the unRAID license for a rainy day.

I was a FreeNAS man before, and a Windows man before that with regards to an OS storage of choice, but unRAID (in my opinion) seems to be the best home storage solution for expandability, flexibility, heaps of tech/forum support, addon's galore, flexible on hardware, budgets, and I can go on and on. I ditched FreeNAS over a year ago as my primary OS storage as unRAID offered much more. unRAID has been rewarding, though a small learning curve to take in, the online support has seen me through any question I had to ask.

 

Good Luck with your new venture, I'm sure you'll be back .

[kenoka]

So basically, it worked so well for you that you hardly had to touch it for four or five years, so you're dumping it? 

[NAS]

I totally see his point.

 

He is at the moment where he needs to refresh his storage and as such has researched if unRAID is still the best fit. Since he is a Apple man several alternate products have more features relevant . The level of storage he needs is very low so the unique unRAID RAID solution is not necessarily required.

 

Add to this he was an early adopter informally promised that his unRAID requirements would be closer met in the near future (multiple times) you cannot blame him for wanting to look about.

 

unRAID is a unique solution and its RAID like clever storage is a killer feature. But its GUI currently sucks, features take an age to come and there is absolutely no visibility of progress/ETA to feature sets.

 

I cannot blame you and if i was in your position i would almost certainly do the same.

[WeeboTech]

I thought DROBO was a great product until I read some of the horror stories.

Maybe it's better today.

 

It does seem like a big point in leaving unRAID is about AFP.

 

Reality is that the hardware is older technology and requires a fair amount of refresh to come up to current day usage.

 

unRAID is a great solution, it may not be the end all of solutions, yet I would suggest being very mindful of the drobo, the state of your data and how much support you will receive in the future.

 

Good Luck!

[prostuff1]

I understand where (s)he is coming from.  I know AFP is coming in 5.0 but when 5.0 is completely ready is anyones guess.  When I sat down and looked through the pros and cons of unRAID (knowing that development was fairly slow) I still chose unRAID.  I would love AFP for the simple fact of backing up via TimeMachine (which you can do via SMB with a few tricks), but native TimeMachine via AFP would be much better.

 

Other than the lack of AFP out of the box I liked everything unRAID gave me.  The ability to tweak and change it to my hearts content was great also.  The forum support here is great and the Add-On's for pre 5.0 are great (Thanks JoeL, WeeboTech, BubbaQ, bjp, etc).  I have been able to do more with my unRAID server than I first expected to and that is wonderful.  When 5.0 is finalized and the event callouts are in place and working we should see some great progress towards easy to use addon's.

[poofyhairguy]

As such I have tested/ran servers for AFP on a lot of platforms, both at work and at home. Because I have this experience, I have some definite opinions about serving AFP, and what the costs are and the advantages/disadvantages.

 

Adding AFP to Unraid would be a big deal because the SMB part of OSX has stagnated since 10.5 and performance between an Unraid box and a OSX box is dismal.

 

With that said I only use OSX desktops and laptops, so for now I just put up with the slow speeds. I can't wait for version 5 though....

[dschur]

The biggest part of my moving is simply due to age of the hardware. I really don't feel safe anymore running 5 year old hardware across the board. I'm worried that if I put any investment into maintaining the array it would be a black hole, start with one disk and then keep going and going and going, with downtime and hassle each time.

 

The motherbaord, memory, hard drives, PATA/SATA controllers (i think there are 4 of them), fans are all 5 years old. I am just going to count my blessings and move onto something else with all new hardware. This meant I was moving to new array regardless.

 

The cost of the hardware/software was less and issue to me than the time. I want something I can take out of the box and have up and running in no time. While unRaid has a lot of benefits, plug and play is definitely not one of them as you are pretty much system building from scratch. It probably took me 3 weeks (elapsed) ordering/installing/configuring parts for my first array, and I wasn't really looking forward to that this time around. My current array is a complete rats nest of cabling/power. You have to admit drobo has the hardware side nailed, nary a screwdriver, just pop the drives in.

 

Yes, AFP did play into it for me. I have a dinky 500GB time capsule now, and the thought of just carving a hunk off Drobo dedicated to it is highly appealing (simple and I get disk redundancy which Time capsule does not have). To be honest AFP on Unraid is a white elephant to me - I have been hearing about it for years now and have yet to see it. I'd be frankly surprised if it's released before this time next year. From what  I've witnessed of recent unraid betas, I would not be willing to use it in beta until the official release.

 

So I was looking at new hardware either way as it's improbable to rebuild in place. By the time I got a new pc case, power supplies, memory, cpu, etc it was probably a wash or close to it with the $650 I blew on a Drobo FS. The drobo has a few other advantages for me - compact size (I will reincorporate it into my office where I can see the status lights - Unraid was just too loud and big in a CM stacker) and email notification (I'm sure I could have done that with unraid as a weekend project).

 

Who knows, maybe I will come back to unraid when I outgrow the drobo, but for now the drobo and drives will be here tomorrow and I'll restart unraid one last time and start pulling my data off after a 5 minute set-up of the drobo. Then, my network will be busy for a bit as  I figure it will take about a week or two to move over 4.5TB. I wanted something simple, clean, fast, and easy to have running right away to start moving data off my unaid before I loose a second drive (a risk I am willing to take for a very short term without replacing the failed drive), so for me, at this time drobo fit the bill. I'm not saying it's fora everyone at all, unraid is great. I do think unraid is probably a better solution for non-Mac users who have very large storage requirements (15TB+) on one array or enthusiasts who like scraping knuckles putting them together, just wasn't for me this time around. I do think both are far better than standard raid arrays for home users.

 

I would equate Unraid to a linux workstation that one builds up from scratch out of leftover parts and has fun tinkering with, drobo to a macbook air, it's sealed, you can't monkey with it, and it has little expandability. I think my own shift from unraid to drobo reflects this - I used o run 4 or 5 windows pc's I had built from scratch and spent a lot of time maintaining 2 years ago. I now have replaced them all with Macs and I never open them up.

 

 

 

 

 

[WeeboTech]

make sure you take the data off the failed drive first.

 

I used o run 4 or 5 windows pc's I had built from scratch and spent a lot of time maintaining 2 years ago. I now have replaced them all with Macs and I never open them up.

 

it's funny, it seems to be the trend these days.

I met a guy in a store one day and he told me the story of when he would always buy Dells. Then max them out with everything, then hardly use all the features.

Then he switched to a mac where he did not have the option to ugprade everything.

He felt it was better to accept what they offered as he would not overbuy and he did not have to concern himself with all the options.

 

I've heard this and "Apples just work" many times.

[dschur]

make sure you take the data off the failed drive first.

 

 

Thanks, good advice - when I restart the array, the first thing I was going to get off the failed drive (/mnt/disk2) was the directory listing. I suspect a lot of it is media files (movies) so they may be a secondary priority to some other contents (photos libraries, documents, etc) in the rest of the array. I was going to move, rather than copy contents off the user share, whittling my way down to zero (and finding some things I probably no longer want along the way).

 

I would replace the failed drive, but given the birdsnest of wires (13 power leads and PATA/SATA cables - most PATA), and dust inside the case, that may be more risky than just running it as-is (I de-cabled the failed drive) for now to get the data off. Just for giggles, when I'm done I'll zero the failed drive and re-init the old unraid array. who knows maybe the thing will keep running and I could use it once every 6 months or so to backup the new array.

[ftp222]

If you are seriously considering Drobo, make sure you test multiple writes/reads at the same time if that is something you intend to do on a regular basis.  I have not tested the FS model, so maybe they are better, but I owned both the Drobo and Drobo S and performance went through the floor when you tried to read/write to the unit at the same time.

 

Also, be sure to visit the drobospace forums - are you still required to have a serial number to view the forums?  Any company that will not open their forums for pre-sales support should not be a strong consideration.

 

Don't get me wrong, I absolutely loved the concept of Drobo and thought it would be the end-all of storage, but that was not the case when i used them.  Between SERIOUS performance issues and near loss of all of my data (among several horror stores on the Drobo forums), I won't go near another one again.

 

Good luck in your hunt for a viable solution!

 

[bkasten]

As such I have tested/ran servers for AFP on a lot of platforms, both at work and at home. Because I have this experience, I have some definite opinions about serving AFP, and what the costs are and the advantages/disadvantages.

 

Adding AFP to Unraid would be a big deal because the SMB part of OSX has stagnated since 10.5 and performance between an Unraid box and a OSX box is dismal.

 

With that said I only use OSX desktops and laptops, so for now I just put up with the slow speeds. I can't wait for version 5 though....

 

I personally would love an easy to install/initiate AFP support. And I would use it.

 

The point I was trying to drive at was I have also had some (very) bad experiences using AFP/Non-Apple combos. There are newer versions of netatalk out since I last tried, so I cannot speak to those, but I have had the older versions of netatalk blow-up on me pretty bad.

 

My desire for a working AFP is for the speed too. Nothing, in my experience, is faster for Macs.

 

My concern is that it may not be the panacea that we Mac users want. I also did not/do not base my needs on whether the protocol is supported when I began my search for a media/file server. I am not a real fan of SMB or NFS either, both are kind of slow, but I have never had them blow-up on me either.

 

In any event, dschur, the original poster expressed his/her disappointment with not having it as an option. If I understand his/her concerns, it is that it was tacitly promised that it would be available, and is not. I am not sure how to address that disappointment.

 

I do know that other company's (Non-Apple) implementation of AFP has been, well, historically bad. After reading the user comments about Drobo for instance, I began to suspect there AFP support, again. I would not like to see unRAID users jump ship for the promise of AFP support from another vendor, and have that lead to tragic results.

 

=======================

 

Skip this part if you are not interested in some of my history supporting AFP in the corporate World. 

 

I once worked for a company that the IT Dept. was pathologically Windows oriented. Anti-Unix, anti-Linux, anti-anything but Windows. Don't even mention Macs. This was a bit of a conundrum however. The company lived by publishing catalogs, and that work is done on Macs. So they allowed Mac users, but Windows Servers. The situation became so bad that they were willing, after some years, of trying to make this work, that they investigated ever other option (well, except Macs of coarse) to fix their issues. After years of testing, and testing, and testing, it was finally agreed that the only server platform that would work for serving AFP would be Mac servers. It was painful for some of these folks. I was just sure that some of these people would rather poke out their own eyes rather than spend IT money on anything Mac, for the AFP support. In the end, they had to do it, and now everyone is much happier.

 

I have never seen AFP done well by anyone but Apple.

 

Even still, if AFP were to be brought to unRAID, I would use it. I just would not base my purchase, or re-purchase on unRAID based on AFP support.

[NAS]

The thing is we are all unRAID fans already.

 

If there was a fancy image on the unRAID homepage that said "Now with added Apple... it just works" I suspect we would have a flurry of "hi I am new here but some apple people told me about this cool product".

 

Features attract users. lacking one key feature to potentially attract loads of apple users.... god knows why it was not added 2 years ago. Anyways I am getting OT

[kizer]

Are we talking the same Drobo here?

 

Man thats some spendy gear.

 

[dschur]

Are we talking the same Drobo here?

 

Man thats some spendy gear.

 

 

Yep, specifically the one I ordered was the FS:

http://www.droboworks.com/Drobo-FS.asp

 

For $650 shipped from Amazon. To be honest it's then less outlay then I originally spent buying a CM stacker case, mobo, cpu, ram, nic, controller cards, cables.

 

Current pre-built unRaid (from lime-tech) is all in the 1200-1400 range. Admittedly most of this stuff holds 15 drives (including my stacker case), but for me my storage needs will be met with 5 bays for a good bit into the future with 2TB drives (or 3TB). Drive size increases have made this possible for me as 500GB was the sweet spot when I started my unraid, and now 2TB are.

[squirrellydw]

make sure you take the data off the failed drive first.

 

I used o run 4 or 5 windows pc's I had built from scratch and spent a lot of time maintaining 2 years ago. I now have replaced them all with Macs and I never open them up.

 

it's funny, it seems to be the trend these days.

I met a guy in a store one day and he told me the story of when he would always buy Dells. Then max them out with everything, then hardly use all the features.

Then he switched to a mac where he did not have the option to ugprade everything.

He felt it was better to accept what they offered as he would not overbuy and he did not have to concern himself with all the options.

 

I've heard this and "Apples just work" many times.

 

I did the same thing about 6 years ago. Built all my pc's and replaced them about every two years and was always running anti virus and spyware programs on them, I don't have to do that now.

 

Macs just work, once you go Mac you won't go back

[Rob_Esc]

 

..given the birdsnest of wires (13 power leads and PATA/SATA cables - most PATA), and dust inside the case...

 

Could you post a snapshot?

[dschur]

Sure, as long as everyone promises not to tell me that the round IDE cables that I've been using for the last 5 years don't work 

 

[kenoka]

Yep, specifically the one I ordered was the FS:

http://www.droboworks.com/Drobo-FS.asp

 

For $650 shipped from Amazon. To be honest it's then less outlay then I originally spent buying a CM stacker case, mobo, cpu, ram, nic, controller cards, cables.

 

Current pre-built unRaid (from lime-tech) is all in the 1200-1400 range. Admittedly most of this stuff holds 15 drives (including my stacker case), but for me my storage needs will be met with 5 bays for a good bit into the future with 2TB drives (or 3TB). Drive size increases have made this possible for me as 500GB was the sweet spot when I started my unraid, and now 2TB are.

 

A good deal of the cost of those prebuilt systems is in the drive cages, and high end Lian Li case. For a budget minded builder, case, mobo, CPU, RAM and power supply for up to six drives is only about $300. Since you already have a license, you'd be out less than half the cost of the Drobo. Add another $150 and you can expand to fourteen drives. If you want a Drobo, I think it's fine, but the argument from cost equivalence doesn't hold water in my opinion.