June 28, 200620 yr The part of this that really gets me steamed is that Tom is still taking people's money and shipping systems (albeit very slowly) but not offering any support for them. If there were was an emergency or a tragedy then he wouldn't have the time to take people's money. I feel cheated. I've got 1000s of dollars worth of hardware, a case that I spent went weeks modding and painting, and it's all ready to go. But I don't see the point in taking the time to configure it if it offers no protection. I might as well just run JBOD or something.
June 28, 200620 yr Well I guess we "early adopters" are up the creek w/o a paddle Given that Unraid is pretty much done, Tom being MIA, what options do we have? I'm currently looking at: 1. NASLite-2 w/ an Areca card, debian based distro- it will support large arrays and most of the above requirements, but will depend on HW XOR cards. No knowing when it will come out though, presumably end of summer. 2. Freenas, FreBSD based distro- it's free, available now, but the performance is something to be desired as is fault tolerance in RAID 5 (much less 6). 3. Cobbled together Ubuntu variant running Raid 4/5+1 or 6 as NAS or SAN (ISCSI). 4. Win2003 w/ an Areca card. 5. The NORCO NAS that will come out in October 6. The rumoured Infrant 8-12 Drive NAS that will come out end of year. What else is out there? is there a Open Source distro that is: 1. Flash based distro 2. Supports Software Raid 4, 5 and or 6 on a large array 3. Web Managed 4. Minimally supports SMB and NFS 5. Cheap What else are people looking for? Update: Found SME Server 7.0; meets many of my requirements, anyone had experience with it? http://contribs.org/modules/phpwiki/index.php/SME7Manual Update 2: Puppy linux via: http://www.personaltechpipeline.com/reviews/showArticle.jhtml?articleId=184400017&pgno=6&queryText= Update 3: this one includes block level access (= iSCSI), not sure how big it is though. http://www.openfiler.com/
June 28, 200620 yr As far as I can tell, none of those options can handle an array of non-identical discs. I'm not ready to give up just yet. We still have to find out for sure if v234 is really defective or if it is just an issue with one person's particular installation. If we can verify that v234 does in fact work for data recover then I am happy with UnRaid as it is. I think it is obvious, however, that no further enhancements are likely to come forward. To me that is not a big deal as I use is as a DVD server and it does it's job well in that area. Was enough of UnRaid code GPL'd to serve as a base for an open source project?
June 28, 200620 yr Well I guess we "early adopters" are up the creek w/o a paddleI wouldn't call myself an "early adopter", I've sat on the fence for at least 10 months since first finding out about this project. I was holding out for the "spin down sata" facility before deciding to go for it on the IDE only variant recently. There's a wealth of difference between early adopters not having all the enhancements, features and fixes that come with each improvement and the product actually perfoming the task it was sold to do. Right from v1.0 there must be parity protection of our data, that's what we paid for. If it doesn't work don't sell it, if it did work but now doesn't then revert to shipping the last known version that did work, but come clean and let us know what we customers who've given Tom our money need to do to be parity protected. This customer service issue has been the difference between seeing UnRaid as money well spent and currently feeling I've been "had". The crazy thing here is that this is a redeemable position, but just ignoring th eissue or hiding doesn't help anyone including Lime Tech. Mark.
June 28, 200620 yr As far as I can tell, none of those options can handle an array of non-identical discs. Infrant's X raid products do it, and well. Was enough of UnRaid code GPL'd to serve as a base for an open source project? This is an excellent question, it can also be addressed to Infrants OS, since its OSS. I believe Infrants OS is compiled on a sparc instruction set ASIC, but a recompile on x86 may be possible, any takers?
June 28, 200620 yr In support of the "self-help" mode, I started work last night on a "shell script" based alternative for some of the functionality currently performed by Tom's web-based management page. I'm pretty certain it will be impossible to duplicate all functionality from a simple shell script, but I'll see how far I can get. My first step is to simply get a status page somewhat similar to the one Tom provided. I managed a start on it last night. It does do nothing so far but to show the current array status (STARTED or NOT) I'll add functionality as I can. Next will be a listing of the drives in the array and their status. After that, a way to start and stop SAMBA and a way to start and stop the unRaid array. My goal is twofold, to get a better understanding of how the unRaid driver is managed, and to provide an alternative tool usable from the telnet prompt. As far as getting the features rharvey is longing for, they must involve either Tom, or someone willing to get involved in a full scale development effort. (Porting the unRaid driver to the newest stable release of Linux, developing new management tools with security in mind, and then addressing issues as they surface) That gets back to the "moral" issue rharvey previously mentioned. The GNU software license encourages the sharing of source code, exactly for the purposes we all want. (Better, faster, slicker applications based on shared ideas and work.) Tom started this by building on the existing "md" (multi-device) Linux Raid driver. A new development effort based on Tom's work is legal and permitted. ( Let's call it "nextRaid" for now. I'll bet at least a few might be interested) As I said earlier, I personally want to support Tom, because I really want to spend my free time watching movies, and not building the next generation version of unRaid. I have not yet set up a true development machine... guess that will eventually occur if Tom doesn't check in here in the next few months. I don't want to undermine his business, but if he does not provide support, odds are he will be getting less and less business anyway. I certainly have no intent of supplying hardware, or software support :-) as he currently does. I (unfortunately) do have the skills and UNIX background to work in a kernel development mode. (almost 20 some odd years ago, I was trained in UNIX internals and Kernel Driver development under AT&T System-5 UNIX. Now, that is light-years distant from today's Linux kernel, but the general concepts and skills are similar) I will do what I need to keep my unRaid server working, even if I have to debug it myself. I'll also provide advice to others on this forum if I think I can help based on what I know. (It may not always be *good* advice, but hey, it's free) I'm very interested if the earlier release of unRaid has the same difficulty in upgrading a disk as described by rharvey, or if it is installation dependent. We still do not know if it is a management utility flaw, or a unRaid driver flaw, or hardware specific, or whatever... We'll have to wait for more folks to try to upgrade their disks under the 324 release and see. Has anyone else upgraded and/or replaced a disk under the 324 release? If we do go forward sharing.. what should "nextRaid" provide as features?? 1. Reliable Raid-like protection of data contents with the ability to mix drive sizes and expand over time. 2. Security, both for lan shared contents AND for its own management tools. (You will need a password to log in via telnet) 3. Performance (concurrent read/write) This may be better with newer Linux kernels. My looking at the source Tom supplied may give some clues, but the performance we see might be exactly the same with a standard raid array on the version of Slackware Linux he used as his base. If so, it might be addressed already in newer releases. (yes, it means the unRaid driver would need to be ported to the newer version of Linux) 4. NTFS filesystem support (even if only in a read-only mode would allow for MUCH faster data migration). 5. SATA drive support. (probably hit exactly the same issues as Tom here, since if the Linux kernel does not support the spin-down and temperature read features, we can't easily either) 6. LVM support (The ability to appear to network appliances as a single tree of media is nice, but there are workarounds, using symbolic links, so all the media is visible to network appliances on a single shared drive, so this can be a lower priority) It might be too limiting in that drives linked together can't be re-sized. At least provide scripts to create directories with links/shortcuts. 7. Independent spin-down of specific disks. (This way, a small drive with the symbolic links and cover-art images can stay spinning all the time.) Then, you could browse your media collection from a network appliance without having to wait for disks to spin up, and the only delay for spin-up would be after a specific media choice is made, after you select it for viewing/listening. 8. Flexibility... Might be nice to be able to config to have some disks NOT part of the nextRaid array, for temporary files and where absolute performance is important. If Tom had that feature (to network share a disk NOT managed by the unRaid driver) we would have been able to isolate if concurrent read/write performance was unRaid related, or network related. Now... increased security and features will mean increased difficulty in setup, and, more likely that increased support will be required in the initial config, even if much better documentation is provided. This is a huge undertaking... (perhaps bigger than any one person could provide) Any other feature requests as we dream together? (Tom... it's OK if you chime in here too) Joe L.
June 28, 200620 yr Joe, I could be very wrong, but I think between all the Open Source projects mentioned above, there is enough [Linux or FreBSD] code [aka ready packages] to actually compile/integrate into something very useable, *without* [much] actual kernel hacking. For the OS, a base small distribution Puppy Linux or Naslite, (or DSL or SLAX or Freenas) is available and rather usuable IMHO. For the RAID, a RAID 4 (w/ or w/o LVM) would be a decent approximation, raid 6 would be even better I think for large arrays. Infrants X Raid code would be the best, but I'm not sure if it's proprietary or Open Source. (e.g., http://wiki.archlinux.org/index.php/Installing_with_Software_RAID_or_LVM , http://cgi.cse.unsw.edu.au/~neilb/mdadm, http://www.infrant.com/download/GPL/linux-infrant-2.4.20.tar.bz2) For web management there are packages ready for inclusion. (e.g., http://www.murga.org/~puppy/viewtopic.php?t=5801&highlight=) A "best of breed" compelation may be had; if there was someone experienced enough, perhaps like yourself that can review/integrate elements. I'm sure many of us here would willingly support either continued OpenSource development of something like Unraid, or a new small comparable distro, even though few of us have the technical skills to technically help. I am not so Linux savvy myself, but I can lend my time (package research, beta testing, project management/documentation if needed) and contribute some $$, as I'm sure others are as well. I believe the first thing to do is get off of Toms forum, because this conversation is obviously not respectful to Tom's affairs- maybe we can reconvene at AVS forum or equivalent and start a thread for brainstorming this? Your post can be the beginning of a requirements survey...
June 28, 200620 yr I believe the first thing to do is get off of Toms forum, because this conversation is obviously not respectful to Tom's affairs- maybe we can reconvene at AVS forum or equivalent and start a thread for brainstorming this? Your post can be the beginning of a requirements survey... I did not indend my post to be disrespectful to Tom in any way, but to instead dangle a challenge in front of him whenever he logs back in and reads this thread. I respect the work he started. I would love for him to be involved in unRaid, nextRaid, or whatever we eventually use. My basic goal is stable, protected storage on my server, nothing more. My spare time is more valuable to me and my family so my first goal is to not have to re-create a new open-source version of unRaid. I do agree though, if an open-source project was initiated it would certainly not be based on Tom's forum . That would be disrespectful (unless Tom wanted to participate and invited continued use). I for one want to see Tom succeed. It is largely up to him at this point. It does appear as if he shipped product as late as June 18th or 19th. Interesting. Joe L.
June 29, 200620 yr I talked to Tom today and he is over nighting me the missing USB flash drive. I checked every where in the box and did not notice anything. I looked in the folder as well. So, Tom IS ALIVE, but he is currently not supporting his product. I feel a bit taken at this point because I am far from a guru like a bunch of you... and I would really like it if Tom would stay on board and make sure this ship doesn't sink. I asked Tom if he would ship me the software before 324 and told him that there is a thread on the MSG board that the 324 is having issues. I did not hear back. I hope to not bother you all to much once I get this thing up and running.... but you guys seem very helpful and I REALLY REALLY appreciate that!!!! Thank you in advance! -Mike
June 29, 200620 yr Hi Mike, When you say you talked to Tom today, do you mean via e-mail? I assume you'd have mentioned this thread to him (and asked if he's coming back) if you actually spoke to him. -PGPfan
June 29, 200620 yr Author I believe the first thing to do is get off of Toms forum, because this conversation is obviously not respectful to Tom's affairs- maybe we can reconvene at AVS forum or equivalent and start a thread for brainstorming this? Your post can be the beginning of a requirements survey... I agree with the new thread over at the AVS forums. To be honest, I say I agree NOT out of respect or disrespect for Tom and his business, I think respect is a two way street and so fare this has been a one way road. the simple fact is that even if he came back strong from this point, I most likely would not continue to pursue this product. I cannot depend upon something like this type of service. Maybe I sound cynical or just plain ticked off. Well I am. I didn't mention this before but I have Tom's cell number. I have called and left 1-2 messages a day for the past 2 weeks. Nothing, Nada, No response at all. It gets to a point where respect, business sense, etc go out the windows and it becomes time to BUCK UP and face the mess that someone created. Tom has chosen to run and not participate at all. From what was mentioned above, it is obvious that he is still alive and not in a COMA etc. So I am getting more angry at this point then anything else. Anyway I think someone who knows ALOT about Linux and the Un-Raid system should start a thread over at AVS and we should move there. The fact is once Tom gets sick of paying for a forum that is bashing his companies customer service and support (which it is and we all think it is warrented at this point) he will just shut it down. Sorry to sound so mad, but hell, I really am!!! BTW GREAT Raid articles in the latest issue of CPU magazine.
June 29, 200620 yr Hi Mike, When you say you talked to Tom today, do you mean via e-mail? I assume you'd have mentioned this thread to him (and asked if he's coming back) if you actually spoke to him. -PGPfan Yes, I spoke to him via email. I told him about the thread and he should take a look and address the problems. Tom over-nighted me two USB flash drives, one with the 324 software on it and then one with the original software. I wish that Tom would keep us up to date. I know that he is alive and I can't believe that Tom is not even returning critical calls that are left on his personal cell phone. It is a shame.... -Mike
June 29, 200620 yr Mike, Can I ask, when you say "So, Tom IS ALIVE, but he is currently not supporting his product." was that what he actually said OR is it your assumption based on his actions, or lack of. Not being critical of you here but if Tom flat out said (via email) that he is not supporting the product then we have our answer and we now know what we need to do. :'( I'm over on AVS now starting a thread there, we should all head in that direction and just keep an eye on this thread. I don't think anymore can be said here and Tom will get the point if he want's it. I will post here no longer in any of the sub-forums, see you all at AVS.
June 29, 200620 yr No, he did not say that he is not supporting the products, I just said that because of this thread that has been going on for a week+ and the fact that Tom has not logged on in quite some time.
June 30, 200620 yr The current state of affairs are disappointing, but I kinda saw this coming. The only real business model here is one of support and that looks to be all but officially gone now... There were many reasons I wanted to make sure the driver source was released (as seen in the GPL thread), primarily for moral and legal reasons, but also because I feared Tom would perform another disappearing act and we'd never get to see the source. Anyway, I spent some time looking at the driver source trying to figure out how it worked and what would be involved making it hardware agnostic. I was able to get it compiled and "working" on my gentoo system, but I never got to the point where I could manage an array with it. The driver is written in a very non-generic way. It's obvious Tom never intended to make the driver open source and it's built to work only with his "proprietary" management tools. For example, you have to load the driver with the list of devices in the array (because there's no persistent store for a config file...) and all the management is done through a /proc interface rather than ioctls. Oh, and for the record, I don't own any UnRaid products. I've never seen any of the management tools, only the /usr/src/linux-2.4.31 tree as was sent to me by another member here. I was also curious of the effort required in porting the UnRaid driver to the 2.6 kernel. Unfortunately, porting the driver to 2.6 would be a massive undertaking that I have neither the skills or time for. My suggestion would be getting the source code in the hands of some real kernel developers and see if any want to support the concept. My guess is that it would be better (more appropriate?) to make "unraid" a device-mapper target. But you might have a hard time keeping the "standard filesystem" feature where you can just pull out an UnRaid drive today and stick it in any Linux box with reiserfs support and read the drive... I dunno. I also have a hard time believing you'll see any performance improvements with the driver in the 2.4 kernel simply because of the single shared I/O queue and scheduler limitations. The 2.6 kernel makes huge improvements in these areas which should provide for better concurrent read/write performance. I'm a HUGE fan of the the UnRaid concept and I'm happy to provide assistance where and when I can on any completely open-source fork of this product, but my time is extremely limited these days. So get busy, Joe -Brian
July 1, 200620 yr Can I say that? I have been monitoring UnRaid on AVS since last year, and after a hard drive crash with complete loss of data, decided I better do something. I love the Unraid idea of not striping, but still getting parity protection. I hadn't seen any new posts on the Official thread on AVS, so figured everything was going smoothly for folks. So I bought 4 400GB Seagate drives and ordered the USB Flash drive last weekend, which showed up a couple days after ordering. Now I see this thread and all kinds of issues with the current release. I can't afford to store 1TB+ of digital images and captured and edited digital video, just to have the system die without any vendor support. Before I buy a mobo and RAM, maybe I would be better to return my uninstalled USB drive and eat the restock fee and shipping. Has anyone tried to return a product to LimeTech for a refund? Geordon
July 1, 200620 yr I'm in the same boat. I'm thinking about returning my USB drive too. Who knows how difficult that will be.
July 3, 200620 yr I'm in the same boat. I'm thinking about returning my USB drive too. Who knows how difficult that will be. I will let you know. I just sent my "unopened" drive back today. I used Visa to purchase, so shouldn't be much of a problem. Geordon
July 3, 200620 yr I'm in the same boat. I'm thinking about returning my USB drive too. Who knows how difficult that will be. I will let you know. I just sent my "unopened" drive back today. I used Visa to purchase, so shouldn't be much of a problem. Geordon Tom wrote the following on AVS when he last surfaced: As far as the "risk" to try UnRaid - well if you just purchase a Flash, there's really not much risk - it comes with a money-back guarantee, the most you'd be out is a couple 39 cent stamps to send it back via 1st class. It was in this post: http://www.avsforum.com/avs-vb/showthread.php?p=7259455&post7259455 Granted, this was several months ago, but I've never read of him giving anybody a difficult time returning product, in fact, until he went silent (not responding to calls or forum posts) I don't think I remember anybody returning product. I expect he'll do the right thing and refund your money. (But let us know if he does not, as that would not be cool) Joe L.
July 3, 200620 yr Author Yes I agree. I have never heard of him giving anyone a hard time about anything....except on going support. It seems that Tom is alive...read the avs forum thread and the newest post.
July 4, 200620 yr A mutiny? ALOT of work has been done porting to 2.6.x linux kernel along with some features and fixes, primarily to implement SATA properly. There are still some remaining issues, so there will probably be one more release based on the 2.4.x kernel. Regarding product support - I will be able to devote more and more time to this forum starting now and ramping up in the next few weeks. Regarding product returns, refunds, etc. - no need to worry about that. We always honor all return/refund requests, although out of hundreds of combined systems/starters/flashes shipped we've only received 2 refund requests - and one of them re-purchased a Flash later!
July 4, 200620 yr In support of the "self-help" mode, I started work last night on a "shell script" based alternative for some of the functionality currently performed by Tom's web-based management page. I'm pretty certain it will be impossible to duplicate all functionality from a simple shell script, but I'll see how far I can get. My first step is to simply get a status page somewhat similar to the one Tom provided. I managed a start on it last night. It does do nothing so far but to show the current array status (STARTED or NOT) I'll add functionality as I can. Next will be a listing of the drives in the array and their status. After that, a way to start and stop SAMBA and a way to start and stop the unRaid array. Below is a preview of what I put together so far. The shell-script to display status is in place, I need to write another set of functions to show the statistics, and a third to perhaps perform some unRaid management functions. Stopping the array is pretty easy, probably not too hard to start it either once I get a lot more familiar with how things work. This way, even if the web-based page locks up, we can still gracefully recover. Here is a screen-shot of the my newest version of the shell-script based utility. It still has only the disk status page, and no ability to manage the unRaid array, but it does report on what is happening when I replace a disk with another in the same slot. And another image as a background parity scan is being performed (Yup...430 minutes is a long time to wait for the next test cycle): Tom, glad to see you back... You were missed and we were concerned for your health. This time, stay involved. We know a good thing and want to keep our servers running smoothly. If you were active, we wouldn't have to figure all this out by ourselves. There is a major bug in the 324 release when upgrading and/or replacing a defective disks. I've described what I found in other posts, but basically the upgrade process fails when it attempts to clear the target drive (unless the drive already has a partition and file-system) and marks the disk as DISABLED. Once marked as disabled, you cannot recover to it. Joe L.
July 4, 200620 yr Author A mutiny? ALOT of work has been done porting to 2.6.x linux kernel along with some features and fixes, primarily to implement SATA properly. There are still some remaining issues, so there will probably be one more release based on the 2.4.x kernel. Regarding product support - I will be able to devote more and more time to this forum starting now and ramping up in the next few weeks. Regarding product returns, refunds, etc. - no need to worry about that. We always honor all return/refund requests, although out of hundreds of combined systems/starters/flashes shipped we've only received 2 refund requests - and one of them re-purchased a Flash later! Tom, I am glad to hear that you are still around and breathing. I am going to try and be real honest and fair with my post. As a business person and a avid customer of techy things, I am almost insulted by the comment of "A Munity?" A Munity is something that happens when the leader is doing things that are not popular, liked, or someone is trying to be devious and take over. We were abandoned. That is the truth. I love the product, it is obvious that alot of us do, but your actions or should I say in-actions are insulting and very unprofessional. I cannot believe for a second that the latest absence is because you were working on a upgrade the whole time. I am dumb, but not stupid. Fool me once..shame on me....fool me twice and shame on you. I could sit here and try and demand a reason for the disappearing act, but to be honest, I don't care. I think that someone who takes someone else's money for a product and then professes to give support and doesn't, should volunteer a reason for something like this that has happened twice. You have NOT responded to emails, voice mails, posts, etc. If you have upgraded the software great....but don't expect most of us to get excited and forget about how we were left behind for the past few months and ignored just because you have a new release. I hope the current release helps the SEVERE situation alot of us are dealing with. The way I see it is this.....we didn't trust you with just giving us a product..we trusted you with the security, stability, and redundancy of OUR personal data.....and you have and continue to put that in jeopardy by ignoring us..... I want to forgive and forget...but I am human...and don't have that ability. I, like so many others, are looking forward to your response.
July 7, 200620 yr As a business person and a avid customer of techy things, I am almost insulted by the comment of "A Munity?" Sorry, man, it was meant with tongue-in-cheek. ... You have NOT responded to emails, voice mails, posts, etc. That's not entirely true. ... I, like so many others, are looking forward to your response. If you want to talk about this in more detail use email or PM. Thread Closed
Archived
This topic is now archived and is closed to further replies.