October 26, 201114 yr axeman - what other alternatives currently offer 3TB support? I'm on 4.7 but am nearing capacity and will need to add another drive soon. I'd like to get a 3TB and reuse my current 2TB parity drive but may end up getting another 2TB in the meantime. I've currently got 15TB of data. Is 5.12a stable enough to upgrade my array to? no clue- i haven't started looking yet. I am hoping against hope that we see something here. I am not ready to abandon ship just yet. Unless of course my array fills overnight. I don't think you'll find much. I looked around quite a bit before deciding on unraid. If you want to avoid striping and have parity there's flexraid and snapraid both of which are one man shows with communities that don't come close to this one. The lack of updates is a bummer but I think unraid is the best thing out there for storing media with drive failure protection and the support on this forum is really good. To add a data point my server has been up for 41 days running 5.0-beta12a on a supermicro X8SIL-F-O with no errors. I don't have any 3TB drives though.
October 27, 201114 yr Yeah. I didn't think anything else supported 3TB. Definitely wouldn't switch even if anything else did. I love my unraid. Would my system run ok with 3TB drives on beta 5.12a? The only add on I have from unmenu is the APC daemon. Everything else is stock. I built a 20 drive budget build about a year ago. System consists of: mobo - Supermicro C2SEE Desktop Board - Intel - Enhanced SpeedStep Technology - Socket T - 1333MHz, 1066MHz, 800MHz FSB - 4GB - DDR3 SDRAM - DDR3-1333/PC3-10600, DDR3-1066/PC3-8500, DDR3-800/PC3-6400 - ATX 2 x SUPERMICRO AOC-SASLP-MV8 PCI Express x4 Low Profile SAS RAID Controller 1 x Intel Celeron 430 Conroe-L 1.8GHz LGA 775 35W Single-Core Processor BX80557430 2GB RAM all in a norco 4220
October 27, 201114 yr Yeah. I didn't think anything else supported 3TB. Definitely wouldn't switch even if anything else did. I love my unraid. Would my system run ok with 3TB drives on beta 5.12a? The only add on I have from unmenu is the APC daemon. Everything else is stock. I built a 20 drive budget build about a year ago. System consists of: mobo - Supermicro C2SEE Desktop Board - Intel - Enhanced SpeedStep Technology - Socket T - 1333MHz, 1066MHz, 800MHz FSB - 4GB - DDR3 SDRAM - DDR3-1333/PC3-10600, DDR3-1066/PC3-8500, DDR3-800/PC3-6400 - ATX 2 x SUPERMICRO AOC-SASLP-MV8 PCI Express x4 Low Profile SAS RAID Controller 1 x Intel Celeron 430 Conroe-L 1.8GHz LGA 775 35W Single-Core Processor BX80557430 2GB RAM all in a norco 4220 there is a slight chance you could have an issue with those saslp-mv8's in Beta 12 and 12a, but B11 should work fine with that build and 3TB drives.
October 27, 201114 yr I am 4.7 and have four 3 TB Hitachi drives and one SUPERMICRO AOC-SASLP-MV8 PCI Express controller waiting to be used. I have 2 TB space left on my array when full I have to face the decision what to do. I read loud and clear that 5b11 will work with that controller but I do recall that at least one person after going from 12a to 11 saw a speed degradation. What has been the speed difference if any between 5b11 & 5b12a?
October 27, 201114 yr Last post from limetech was over 1 month ago. Has development stopped? I think most users would love at least 1 post stating what is the status of this software. EzBox limetech was online earlier today, so it is most likely that development has not stopped. The last post may have been a month ago but be sure to look at the last activity so see get a true idea of how often LimeTech is on the forum. This statement has been said numerous times, but the same doubt remains. Can the global mods not convince Tom to maybe have a thread that can be updated only by him ... sort of like a blog and if he only types one line a week at least we can see (hear) that development is being done. I work in IT Support and i instruct my guys to report to users even if there is nothing fixed at least they realize they have not been forgotten and surely this cult following deserves this at the very least. This will make a huge difference to people doubts ...
October 27, 201114 yr I am 4.7 and have four 3 TB Hitachi drives and one SUPERMICRO AOC-SASLP-MV8 PCI Express controller waiting to be used. I have 2 TB space left on my array when full I have to face the decision what to do. I read loud and clear that 5b11 will work with that controller but I do recall that at least one person after going from 12a to 11 saw a speed degradation. What has been the speed difference if any between 5b11 & 5b12a? Do you have 2 unRAID keys? I was in close to the same situation—worried that the SASLP-MV8 would give me problems like it has others. I did a fresh install of 5.0b12 using my second key and it worked without issue. I've since performed 2 parity checks, a parity drive swap, a data drive addition, and another parity check with zero array/drive related errors. I don't pay much attention to the parity check speed, but it's usually in the 40-80MB/s range when I check on it. This is with all "green" SATA2/3 drives. YMMV
October 27, 201114 yr I don't want to push limetech to do updates on a weekly basis or in other ways interfere with his decisions to comment or communicate, BUT... This quote below keeps me checking if/when there is a version 5 coming that is recommended for production. Currently the choice is either to stay with a system that has a serious known issue that - as I read it - could lead to data loss (the term "Critical" don't sit well in my stomach) upgrade to a version that is not recommended for production I'll probably end up attempting to upgrade to a 5-version as I upgrade my Hardware, and hope not to have any issues. But IMO it is a valid point from the more impatient in here to either get an update attempting to adress the issues, or maybe limetech/someone looking at all the fragmented reporting of observations and make a summary post like: "This is what you get and these are the known issues for 5.0.b12a". Preferrably from limetech, which would also indicate what the issues are that are being considered, and their severity. At least that would be a help and indicate where the stumbling blocks are (including the kernel-related out-of-limetech-control ones). Hey - maybe someone could even offer dedicated help towards specific issues if this was known (as opposed to guessed upon). Anyway, this keeps me checking Announcements daily, hoping for news of progress. It was certainly nicer when limetech were posting more frequently... There's a critical bug fix in this release having to do with data rebuild. There is a corner case that comes up where, if during a data rebuild of a disabled disk (or disk replaced with a larger one), if a write-request occurs for another disk in the same stripe for the disk currently being rebuilt, it's possible the data for the disk being rebuilt is not actually written. Later, this will cause a Parity Check 'sync error'. If you are running any previous 5.0-beta release, please upgrade to this release. In addition, [red]I will be updating the 4.7 series (to 4.7.1) with the same parity sync bug fix[/red]. How do we avoid this bug? Not write anything while the replacement drive is being rebuilt? By updating to 5.0-beta8 or [red]4.7.1 (which isn't out yet)[/red]. Barring that, yes, with a data rebuild in progress, don't write. With a "upgrade existing disk to larger one" operation, it's potentially unavoidable. Something interesting about this bug: it has existed since day 1 Two months later, and still no 4.7.1. Any idea how much longer for this fix? "Unavoidable" data corruption, if using unRaid's "stable" release... that doesn't look good.
October 27, 201114 yr Perhaps while we wait for the next beta and stable release, we can help each other out. There are hardware configurations that works fine in v5b12a and others that clearly don't. I've started a thread in the Version 5 Development>Other forum to help figure out which are which. Please help if you have the time. http://lime-technology.com/forum/index.php?topic=16099.0 Thanks.
October 27, 201114 yr Perhaps while we wait for the next beta and stable release, we can help each other out. There are hardware configurations that works fine in v5b12a and others that clearly don't. I've started a thread in the Version 5 Development>Other forum to help figure out which are which. Please help if you have the time. http://lime-technology.com/forum/index.php?topic=16099.0 Thanks. This is a good idea but in order to be useful one need to provide the the motherboard BIOS revision and the controller(s) firmware revisions. Second - the observed speed for parity check is not accurate, but once it is done there is an entry in the syslog with the time in seconds and from there one can calculate the accurate speed. Third - UPS is very important for any serious Unraid user as the capacities of the HDs are going up and as a consequence the time for performing array procedures and preclearing of the new HDs is increasing substantially
October 27, 201114 yr I am 4.7 and have four 3 TB Hitachi drives and one SUPERMICRO AOC-SASLP-MV8 PCI Express controller waiting to be used. I have 2 TB space left on my array when full I have to face the decision what to do. I read loud and clear that 5b11 will work with that controller but I do recall that at least one person after going from 12a to 11 saw a speed degradation. What has been the speed difference if any between 5b11 & 5b12a? Do you have 2 unRAID keys? I was in close to the same situation—worried that the SASLP-MV8 would give me problems like it has others. I did a fresh install of 5.0b12 using my second key and it worked without issue. I've since performed 2 parity checks, a parity drive swap, a data drive addition, and another parity check with zero array/drive related errors. I don't pay much attention to the parity check speed, but it's usually in the 40-80MB/s range when I check on it. This is with all "green" SATA2/3 drives. YMMV No I don't have 2 UnRaid keys so that is out of the question, I was considering to get an additional Plus key to build one of the HP-Micro servers with 6 drives, and then do a 5B12a install but I am waiting for the new HP micro server boxes to go on sale, got 4 Hitachi drives $110, what an excellent timing. However I notice that with two servers I believe there is no way to get under a "Movies" share both servers listed. This means I would have to shuffle my content around and basically have one server for TV shows and another for Movies. I thought that this unified share thingy maybe on the road map. You said that you did a virgin install of 5B12a, but I read that many had problems and had to go back to upgrade from 4.7 to 5B12a in order to get a somewhat stable setup. I wish that when people post there success or failure stories include hardware details and Firmware / Software details to make an intelligent guess of what to decide on. In any occasion I still have time but feel the frustration of many of the users here.
October 27, 201114 yr This is a good idea but in order to be useful one need to provide the the motherboard BIOS revision and the controller(s) firmware revisions. Good idea. I'll add that info. Second - the observed speed for parity check is not accurate, but once it is done there is an entry in the syslog with the time in seconds and from there one can calculate the accurate speed. Hmm, do you know a search string I could use to find this info in my syslog? EDIT: Nevermind, I found it. "kernel: md: sync done. time = 33577sec" That's about 9hrs & 20 mins. I have a 3TB parity drive, but my largest data drive is 2TB, so 2TB / 33577s = 60MB/s. Third - UPS is very important for any serious Unraid user as the capacities of the HDs are going up and as a consequence the time for performing array procedures and preclearing of the new HDs is increasing substantially. I agree with you there. I've got a 1500VA for my unRAID sytem. In my tests this is good for ~ 2 hours (@120-200W load) and safely shuts down the array. Are there compatibility problems with certain UPS models or are you getting at something else? Should I add UPS model info to the survey?
October 28, 201114 yr Beta-13 is out!! http://lime-technology.com/download/doc_details/29-unraid-server-version-50-beta13-aio
October 28, 201114 yr piping in... After testing 3-drives, I purchased the PRO 2-Pack in June, 2010. I've been migrating all my backups and media storage to UNRAID since. Current spec should be in my sig. I'm eager to put in 3TB drives. I need 5.x to do it. And I've been waiting for the "stable" or "release" version for a while. Migrating to some other solution is not on my "to do" list at all. But it's starting to irritate me that I'm still buying 2TB drives as the 3TB drives become more available and less expensive. c'mon... I'm just sayin.
October 28, 201114 yr Beta-13 is out!! http://lime-technology.com/download/doc_details/29-unraid-server-version-50-beta13-aio YAY for Tom
November 1, 201114 yr Hi. Very happy R5b12a user, but first post, as unlike the other queries I've had, I am unable to find a definative answer across the rest of the forum. Have unRAID running on a HP ProLiant MicroServer N36L. There is information around reporting that this does not support S3 sleep (the lack of BIOS options, and information on avforums.com would support this) however, my initial trial on 4.7 had the S3.sh script working perfectly. I have taken note of the kernel change, and the differences in how sleep is triggered; "echo 3 > /proc/acpi/sleep" on 4.7 and "echo -n mem >/sys/power/state" on 5.12a. Currently /sys/power/state has no entries; if my understanding is correct, I should see the available sleep states here, but I am assuming this is populated by whatever the BIOS reports (?). So, considering my hardware is unchanged, how can I utilise /proc/acpi/sleep on 5.12a? My thoughts were perhaps that I could turn ACPI off and APM on with kernel parameter options, but unsure of the impact, or whether this is even viable. Any thoughts or assistance would be much appreciated.
November 2, 201114 yr I've been running 4.7 for a while without any problems, but needed to add some more disks to the array and start using my BR10i (LSI SAS1068E). I've been able to preclear 5 different drives, simulate a drive failure and then rebuild without problems running beta12a.
November 6, 201114 yr unRAID decdied it doesnt want to start my array automatically now.. Any ideas? It works fine if I tell it to start by going to the web interface and pressing the start button. Config is valid.. I'm out of ideas. I think it started doing this after I added a drive to the array. Parity was then checked a week after that (on the 1st) and checked out fine with 0 errors. syslog.txt
November 12, 201114 yr unRAID decdied it doesnt want to start my array automatically now.. Any ideas? It works fine if I tell it to start by going to the web interface and pressing the start button. Config is valid.. I'm out of ideas. I think it started doing this after I added a drive to the array. Parity was then checked a week after that (on the 1st) and checked out fine with 0 errors. There's an option in the beta to tell unRAID to automatically start or not. I don't remember exactly where it is, but look around and you should find it. By default the option is set to not start the array.
November 13, 201114 yr Hi all, Last week I tried to set one of my shares to use my cache drive. After I hit "Apply" in the web interface, emhttp hung. I killed it from the console, and tried to start it over, but I'm now getting a segfault. Syslog says: Nov 7 10:29:45 mimer emhttp: unRAID System Management Utility version 5.0-beta12a (Lime Tech) Nov 7 10:29:45 mimer emhttp: Copyright (C) 2005-2011, Lime Technology, LLC (Lime Tech) Nov 7 10:29:45 mimer emhttp: Pro key detected, GUID: 05DC-A731-PBBV-BYYBMTQ7F4A8 (Other emhttp) Nov 7 10:29:45 mimer emhttp: rdevName.22 not found (Other emhttp) Nov 7 10:29:45 mimer emhttp: diskFsStatus.1 not found (Other emhttp) Nov 7 10:29:45 mimer kernel: emhttp[20127]: segfault at 0 ip b74a1760 sp bfcac0a0 error 4 in libc-2.11.1.so[b7428000+15c000] (Errors) A forum search shows that this error has been seen as early as 5b8, and also at least for one person also running 5b12a. Any ideas on how to proceed? Attached is the output of mdcmd status. mdcmd_status.txt
November 26, 201114 yr User of 5.12a with no problems since launch on a N36L Proliant Microserver. Backing up my desktop PC to the unraid slowed down however and I noticed these error lines in the syslog. The speed subsequently increased again. Nov 24 20:37:40 shortie kernel: [<c105f857>] warn_alloc_failed+0xb2/0xc4 Nov 24 20:37:40 shortie kernel: [<c105ffc2>] __alloc_pages_nodemask+0x456/0x47f Nov 24 20:37:40 shortie kernel: [<c128fb68>] ? skb_copy+0x2e/0x83 Nov 24 20:37:40 shortie kernel: [<c106003f>] __get_free_pages+0xf/0x21 Nov 24 20:37:40 shortie kernel: [<c107e11a>] __kmalloc_track_caller+0x2c/0xee Nov 24 20:37:40 shortie kernel: [<c128f0ee>] __alloc_skb+0x53/0xf1 Nov 24 20:37:40 shortie kernel: [<c128fb68>] skb_copy+0x2e/0x83 Nov 24 20:37:40 shortie kernel: [<f84463af>] tigon3_dma_hwbug_workaround+0x36/0x1b1 [tg3] Nov 24 20:37:40 shortie kernel: [<f8446bed>] tg3_start_xmit+0x6c3/0x7a8 [tg3] Nov 24 20:37:40 shortie kernel: [<c1295bad>] dev_hard_start_xmit+0x245/0x31c Nov 24 20:37:40 shortie kernel: [<c12a4466>] sch_direct_xmit+0x50/0x137 Nov 24 20:37:40 shortie kernel: [<c1295d82>] dev_queue_xmit+0xfe/0x274 Nov 24 20:37:40 shortie kernel: [<c12b0127>] ip_finish_output+0x227/0x262 Nov 24 20:37:40 shortie kernel: [<c12b01c6>] ip_output+0x64/0x68 Nov 24 20:37:40 shortie kernel: [<c12adf4d>] ip_local_out+0x57/0x5b Nov 24 20:37:40 shortie kernel: [<c12afb5a>] ip_queue_xmit+0x2a5/0x2f2 Nov 24 20:37:40 shortie kernel: [<c12beee8>] tcp_transmit_skb+0x4d7/0x50d Nov 24 20:37:40 shortie kernel: [<c12c11f7>] tcp_write_xmit+0x2f9/0x3d7 Nov 24 20:37:40 shortie kernel: [<c12c1319>] __tcp_push_pending_frames+0x18/0x6f Nov 24 20:37:40 shortie kernel: [<c12bdfc9>] tcp_rcv_established+0x501/0x578 Nov 24 20:37:40 shortie kernel: [<c12c33b5>] tcp_v4_do_rcv+0x46/0x137 Nov 24 20:37:40 shortie kernel: [<c12c386c>] tcp_v4_rcv+0x3c6/0x647 Nov 24 20:37:40 shortie kernel: [<f847935b>] ? ahci_qc_prep+0xb3/0x126 [libahci] Nov 24 20:37:40 shortie kernel: [<c12ac081>] ip_local_deliver_finish+0x93/0x158 Nov 24 20:37:40 shortie kernel: [<c12ac172>] ip_local_deliver+0x2c/0x2f Nov 24 20:37:40 shortie kernel: [<c12abd8b>] ip_rcv_finish+0x263/0x28b Nov 24 20:37:40 shortie kernel: [<c12abfbb>] ip_rcv+0x208/0x23b Nov 24 20:37:40 shortie kernel: [<c1293604>] __netif_receive_skb+0x234/0x25a Nov 24 20:37:40 shortie kernel: [<c1294a5f>] netif_receive_skb+0x5d/0x63 Nov 24 20:37:40 shortie kernel: [<c1294b1c>] napi_skb_finish+0x1e/0x34 Nov 24 20:37:40 shortie kernel: [<c1294f4c>] napi_gro_receive+0xc7/0xcf Nov 24 20:37:40 shortie kernel: [<f84484bf>] tg3_rx+0x3f6/0x548 [tg3] Nov 24 20:37:40 shortie kernel: [<f8448668>] tg3_poll_work+0x57/0x112 [tg3] Nov 24 20:37:40 shortie kernel: [<c1003bb6>] ? handle_irq+0xd/0x6a Nov 24 20:37:40 shortie kernel: [<f844887f>] tg3_poll+0xae/0x1d8 [tg3] Nov 24 20:37:40 shortie kernel: [<c1295022>] net_rx_action+0x59/0x12a Nov 24 20:37:40 shortie kernel: [<c102ccce>] __do_softirq+0x6b/0xe5 Nov 24 20:37:40 shortie kernel: [<c102cc63>] ? irq_enter+0x3c/0x3c Nov 24 20:37:40 shortie kernel: <IRQ> [<c102cb21>] ? irq_exit+0x32/0x53 Nov 24 20:37:40 shortie kernel: [<c100360b>] ? do_IRQ+0x7c/0x90 Nov 24 20:37:40 shortie kernel: [<c130b8a9>] ? common_interrupt+0x29/0x30 Nov 24 20:37:40 shortie kernel: [<c1007e14>] ? default_idle+0x2e/0x43 Nov 24 20:37:40 shortie kernel: [<c1007f3f>] ? amd_e400_idle+0xb1/0xcb Nov 24 20:37:40 shortie kernel: [<c1001a60>] ? cpu_idle+0x3a/0x52 Nov 24 20:37:40 shortie kernel: [<c1306753>] ? start_secondary+0xad/0xaf I don't know if they're relevant but posting here anyway. Full syslog attached. The error occurs a few times. Cheers Eric syslog-2011-11-251.txt
December 3, 201114 yr Bug Report, When accessing the web interface on 5b12a from my computer unraid seems to tell the drives to spin up. Full stats as well as syslog are in this post http://lime-technology.com/forum/index.php?topic=17040.0 JoeL. suggested this report be made. Bug is found with and without having simple features installed.
December 4, 201114 yr Interesting, I thought I was seeing this same thing for a while using beta12a, but my problem seems to have gone away. It would take longer than usual to open the main unRaid web page the first time after authenticating, and then all the drives would be spun up. Unfortunately I don't recall doing anything to resolve the issue and I can't reproduce it now.
December 4, 201114 yr It's been happening probably for each beta but I know for a fact that it's been happening since beta10 and still happens on beta 13. Been talked about every so often. Every now and then when you visit the web admin, the drives will spin up one by one and the web admin won't respond until they all spin up. The web admin will show the drives as being spun down but if you refresh the page, they are shown spun up. A couple of times for me, it's spun up every other drive instead of all. Last time it happened, I pulled up the unmenu port and it reported the drives spinning up. I would refresh and unmenu would pause until the drive spun up, then it would return and show the next drive was spun up. Rinse & repeat. Unmenu isn't (likely) the cause, I didn't start running it until beta12a and the issue happened on three prior betas.
December 13, 201114 yr Running the cache_dirs script is what made mine quit spinning up all drives when opening the initial web page. If it's not running the web page takes 20 seconds to load, and all the drives are spun up. If it is running, it opens immediately without spinning everything up. I am using a LSI (IBM BR10i) controller if that matters.
December 15, 201114 yr Hi, Several weeks ago I moved from 4.7 to unRaid 12a. Shortly after I also moved onto a completely virtualised server using passthrough to access my disks from an LSI card. I was about to replace my cache drive with a new SSD, and I suddenly discovered that most of my recent TV shows are still sitting on my cache disk. Looking in the logs, I discovered that my disk1 share, where all TV Shows had been copying until now, is full. Also the mover script has been failing for weeks due to the full disk. It's been a while since I set up my user shares, so I had to quickly revise how things work in the wiki. I have checked the settings - it is set to a min free space of 5 GB and an allocation method of High Water, with split level set to 0, with no include or exclude disks. I tried manually creating a TVShows disk share on disk 2 (which has plenty of empty space) but when I run the mover script from the web GUI, it always appears to be trying to copy content to disk1 and fails. I've tried changing to "Most free" with no change, also added an include disk for "disk2" also with no luck. I tried changing the min free space, but it won't "stick" when I apply the change. It always reverts to High water. Syslog attached. Can anyone please help? I think the settings are correct, so I'm wondering if it is a beta issue? Note I also checked the user permissions and ownsership on the new TVShows disk share on disk 2 that I created, and it looks the same as disk 1. Thanks syslog.txt
Archived
This topic is now archived and is closed to further replies.