flambot

Woke up this morning and the parity check was complete - WITHOUT any sync errors. This surprised me - that something is obviously wrong with disk15 and there weren't any errors. I was expecting there to be. Another thing that is surprising. The log for this session isn't on the flash drive. The last thing the log says under syslog (unmenu) is "Jan 11 04:40:01 Tower syslogd 1.4.1: restart." - with only a few short entries - all about disk spindown as the smaller ones finished parity check. I guess the log got full?? My did the errors stop when I shutoff cache_dirs? Better go and do the file system check. Thx for the help.

Greetings, I am currently running a parity check (to check the parity sync) after removing a drive. I decided to check the syslog using unMenu and discovered a whole lot of errors in it - all the same as this sequence (repeating). It started immediately upon startup this morning and continued all day. I just noticed this after the parity check has been running approx 9 hours. There are NO sync errors showing on the WebGUI. Fearful I would use up all the RAM, I turned Cache_dirs off and this error immediately stopped. Not sure what I should do now?? The relevant part of the syslog is attached. Jan 10 10:08:00 Tower cache_dirs: cache_dirs process ID 2409 started, To terminate it, type: cache_dirs -q Jan 10 10:09:55 Tower kernel: attempt to access beyond end of device Jan 10 10:09:55 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 10:09:55 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1299 0x0 SD] Jan 10 10:09:55 Tower kernel: REISERFS (device md15): Remounting filesystem read-only Jan 10 10:09:55 Tower kernel: attempt to access beyond end of device Jan 10 10:09:55 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 10:09:55 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1300 0x0 SD] Jan 10 10:10:25 Tower kernel: ata11.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen Jan 10 10:10:25 Tower kernel: ata11.00: failed command: READ DMA EXT Jan 10 10:10:25 Tower kernel: ata11.00: cmd 25/00:08:37:89:e5/00:00:1d:00:00/e0 tag 0 dma 4096 in Jan 10 10:10:25 Tower kernel: res 40/00:ff:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout) Jan 10 10:10:25 Tower kernel: ata11.00: status: { DRDY } Jan 10 10:10:25 Tower kernel: ata11: hard resetting link Jan 10 10:10:26 Tower kernel: ata11: SATA link up 3.0 Gbps (SStatus 123 SControl 300) Jan 10 10:10:26 Tower kernel: ata11.00: configured for UDMA/133 Jan 10 10:10:26 Tower kernel: ata11.00: device reported invalid CHS sector 0 Jan 10 10:10:26 Tower kernel: ata11: EH complete Jan 10 10:10:26 Tower kernel: attempt to access beyond end of device Jan 10 10:10:26 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 10:10:26 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1301 0x0 SD] Jan 10 10:10:40 Tower kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 52428 does not match to the expected one 1 Jan 10 10:10:40 Tower kernel: REISERFS error (device md15): vs-5150 search_by_key: invalid format found in block 176063846. Fsck? Jan 10 10:10:40 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1380 1248 0x0 SD] Jan 10 10:11:54 Tower kernel: attempt to access beyond end of device Jan 10 10:11:54 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 10:11:54 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1299 0x0 SD] Jan 10 10:11:54 Tower kernel: attempt to access beyond end of device Jan 10 10:11:54 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 10:11:54 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1300 0x0 SD] Jan 10 10:11:54 Tower kernel: attempt to access beyond end of device Jan 10 10:11:54 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 10:11:54 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1301 0x0 SD] Jan 10 11:12:49 Tower kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 52428 does not match to the expected one 1 Jan 10 11:12:49 Tower kernel: REISERFS error (device md15): vs-5150 search_by_key: invalid format found in block 176063846. Fsck? Jan 10 11:12:49 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1380 1248 0x0 SD] Jan 10 11:12:54 Tower kernel: attempt to access beyond end of device Jan 10 11:12:54 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 11:12:54 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1299 0x0 SD] Jan 10 11:12:54 Tower kernel: attempt to access beyond end of device Jan 10 11:12:54 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 11:12:54 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1300 0x0 SD] Jan 10 11:12:54 Tower kernel: attempt to access beyond end of device Jan 10 11:12:54 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 11:12:54 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1301 Jan 10 19:00:01 Tower kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 52428 does not match to the expected one 1 Jan 10 19:00:01 Tower kernel: REISERFS error (device md15): vs-5150 search_by_key: invalid format found in block 176063846. Fsck? Jan 10 19:00:01 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1380 1248 0x0 SD] Jan 10 19:00:05 Tower kernel: attempt to access beyond end of device Jan 10 19:00:05 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 19:00:05 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1299 0x0 SD] Jan 10 19:00:05 Tower kernel: attempt to access beyond end of device Jan 10 19:00:05 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 19:00:05 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1300 0x0 SD] Jan 10 19:00:05 Tower kernel: attempt to access beyond end of device Jan 10 19:00:05 Tower kernel: md15: rw=0, want=27487790696, limit=1953525104 Jan 10 19:00:05 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1294 1301 0x0 SD] Jan 10 19:00:05 Tower kernel: REISERFS warning: reiserfs-5090 is_tree_node: node level 52428 does not match to the expected one 1 Jan 10 19:00:05 Tower kernel: REISERFS error (device md15): vs-5150 search_by_key: invalid format found in block 176063846. Fsck? Jan 10 19:00:05 Tower kernel: REISERFS error (device md15): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [1380 1248 0x0 SD] Jan 10 19:00:08 Tower cache_dirs: killing cache_dirs process 2409

Hi Joe, Cache-dirs is about the only thing I run on my server apart of the APC UPS Powerdown script. I have only recently added it after upgrading to 4.5 series from 4.3.3. Is there any time I should turn it off - like when I'm copying a lot to the server or running a parity check? I have 2Gb RAM and it seems to handle everything so far - but a couple of crashes (can't connect to the server) have started to plague me since I upgraded to the 4.5 series and I'm wondering if it running in the background while I do other tasks is having an affect? You did comment in another post how it can use all the RAM, but is there a situation it is better to turn it off? Also...If I stop it via the "-q" do I start it again with "-W" Does this make it start again with all the features as in the "Go Script"? And...as part of my trouble shooting, I'm considering downgrading, so I am assuming cache_dirs will work with 4.3.3, but that the double stop will take effect? Thx.

Me too.... Mine has run perfectly for 3 years, then after upgrading to 4.5.6 I started having issues. Something happens and you can't connect to the server using any means. I'm assuming I'm having a similar problem where the server has crashed, but am unable to grab a log also. Thx Tom for the log advice.

Greetings, I wonder if some of you skilled Linux Guru's could help me out please. I've recently built a small Ion-based media player (HTPC) to attach to my unRAID server (and replace my old Xbox). It is running xbmc using a minimal Ubuntu install called LIVE (Based on Lucid I think) - that is actually installed on a SSD HDD (32Gb). I have added a 2nd HDD to the case (a SG 500Gb 3.5" one). I've read and researched numerous guides and posts about setting this all up. I have managed to get it working (more fluke than skill) and I'm really worried I haven't done it right and this will impact me at some point. The 2nd drive is for music/pics media (Partition 2), and partition 1 for a backup of the SSD so I can backup the complete install. I've since decided I'd like to add another partition (3 in total) so I can load more recent versions of the Live/xbmc s/ware to test and iron out all the issues before committing to it. I'm not sure if it is possible to have the backup partition or the test version partition work as I've stated?? Here is how I went about it. I used "cfdisk" to create 2 partitions (I want to change this now to 3 partitions). They were both primary partitions - one 32Gb in size and the other the remaining portion of the disk (about 468 Gb). Using "fdisk -l" I discovered the SSD was labelled sda1,2,3 (3xpartitions I think) etc and the 2nd HDD was called sdb (sdb1 for the backup partition, and sdb2 for the media partition). Once these were created, I tried to use the command "mkfs.ext3 -b 4096 [sdb1]" to format the partitions, but it gave some message that didn't mean much. I ended up using this one "mkfs -t ext4 /dev/sdb1" and "mkfs -t ext4 /dev/sdb2" and that seemed to work (not sure this was right?). I moved on to mounting them then. This part was really confusing as I don't understand why you have a mount point (seemed to be under a folder??) and why it just doesn't show up like a drive does in windows. (WHY??) My reading said I should set the drive up under the /media folder in the SSD's root, so I used "sudo mount /dev/sda1 /media/sdb1". Later I found these folders under /media - "7c011d8d-7ed6-4cc5-868d-92e35ec6178e" and "c9ea0b4b-8f75-453d-bc31-42b2d9bd38b4" Not sure why the long letter/number name but I figured it was the partitions. See here - Can I re-name these long names? I'm betting they will change when I alter things to add the 3rd partition. Should they even be here? Anyway, I found the partitions did mount, but that they didn't survive a reboot. I read I had to edit the fstab file, but in my playing I did something that made them survive a reboot - without modifying fstab. (Sorry if some of this isn't clear - but I've been playing with this for several days and my memory is less than perfect, even though I tried to keep a written record of what I did). I did play with these commands - "mount /mnt/sdb1" and also this one "mount | grep /dev/sdb1" and "mount | grep /dev/sdb1 /media/ 7c011d8d-7ed6-4cc5-868d-92e35ec6178e" etc. but not sure if either of these made it survive a reboot. After all this, I found I couldn't write to the partitions - and read I had to set the permissions. Wow...this linux is tough I found these commands in a post to set the permissions. Definitely not sure what they do, but it seemed to work. sudo chown -R xbmc: /media/c9ea0b4b-8f75-453d-bc31-42b2d9bd38b4 sudo chmod -R 777 /media/c9ea0b4b-8f75-453d-bc31-42b2d9bd38b4 I need some help to understand if what I have done is correct, or if it'll create further troubles done the track. At present, the Ionbox is in a testing phase before I take it into the lounge for everyday use. And...two more things. 1 - Is there an easy method to back up the SSD to the newly created partition (and then restore to the SSD should I need to)? I have read about something called clonezilla but I was wondering if this can be achieved from the terminal window?? 2 - Can the 3rd partition I intend to create be used for a testing area for a "TEST" complete install (O/s and xbmc) - meaning - something I can start using the file manager within xbmc and then exit out of it to return to the usual install? (I can do this with the xbox currently, but have no idea if it's possible under linux. Sorry for the long post and I appreciate any help that can be offered.

I turn mine off too, but WOL sounds nice. Not at the risk of corrupted data though. Does your XBMC comment mean you turn your unRAID Server off with XBMC? If so, how? Is there a post that explains a cron job for shutdown after a certain time if inactive? That sound interesting. This sounds great. Not so sure about this though. My server has been running very well for 3.5 years. Don't want to jinx it

Thx for the link to the review. Those are nice. I found some locally here in NZ, but they cost $260 each - too expensive to fit to my 16HDD case. I have my HDD's in Thermaltake icages - 3 HDD's per cage with a 120mm fan. Keeps the drives below 30 degrees usually. They are a PITA though to replace/upgrade the drives. I think next time I'm going to build my unRAID into a cupboard so I can easily get to everything.

While a little OT compared to the initial post, have any of you encountered data corruption problems using S3? I think I read a post from Weebotech (maybe BubbaQ) where they made a comment that it's not such a good idea using standby on a raid system?? I'd like to because power costs just keep going through the roof, but things like this make me uncertain how to proceed.

Hey Chris Pollard - That case is cool. Do you have a pic of how those icydock things open. I guess the drive comes out the front? Are the drives easy to get in and out? I'm assuming the drive plugs into what you call a backplane? With 5 drives in that top bay, does that mean you have 5 SATA cables plugging into the back of it? (Doesn't look like that many). Thx

I managed to recover the files I wanted by using "Recuva" - by the makers of CCleaner. Seems to be a great piece of free software.

I use a Hard Drive Dock with several HDD's. One is suddenly inaccessible. A box pops up saying "The directory or file is inaccessible or corrupted". I have used the Sea Tools software and carried out a test that reveals the HDD is still healthy. XP's "Computer Management" area also says the same thing. What do I do now? Can I recover the info on the HDD? Or, can I see what is on the drive to see if it's worth recovering. I know many of the files also reside on my unRAID server so I can restore them - but am not sure what else I may have put on the drive. Is must be something that states what's in the root of the drive. This must be corrupted. Currently I am performing a longer test and so far no bad sectors have been found. Not sure where to go from here? Thx

Thx WeeboTech - looks like it's just what I needed. I'll give it go once I get my upgrade out of the way.

Not sure where to post this?? I've been a long time user of unRAID for my video library. I use an XBox running XBMC to access it. After an upgrade of XBMC to a later version, I noticed that unRAID was spinning up all the movies drives (XBMC is set to load straight in the movie library at turn on). Previously, I could browse my library WITHOUT the NAS on (as everything is loaded into a SQL library and cached locally), but now, browsing will spinup the HDD"s if the NAS is on. I could also switch from the movie side on XBMC to the TV Side without any HDD"s spinning up - now all the TV HDD"s spinup when I make the switch - OR - if the NAS is off, then switching to the TV side takes forever to load (when previously it was almost instantaneous). It is like the XBOX is now looking for something on the NAS (to verify something??), but local caching should negate this shouldn't it? I have been trying to find the reason for this for a long time. I have posted on XBMC forum several times without success. I have also tried heaps of later releases of the XBMC software and I still suffer this issue. Then I realised I upgraded from 4.1 to 4.3.3 round about the same time as I upgraded to the later version of XBMC. Could unRAID be responsible for spinning the HDD"s up independently of the XBMC software? Everything in my media library is cached at the Xbox - so there is NO need to spin-up the drives. SO how do I find out what is going on? I have also read something about a function - I think it caches the dir structure locally (On the unRAID box) - is it "IS" or "IR" or something. It supposedly refreshes automatically using a fairly short interval. Will this help alleviate this problem. I realise a log will help, but I thought if I could get some pointers on what to look for I could start making some in-roads. I have also noticed a tendency for the HDD"s on the unRAID box to sometimes stay spun up if I leave the Xbox on (which I've been doing to stop the HDD"s spinning up when I turn the xbox on). This issue seems related somehow as well - but it does NOT always happen. Thx for the help.

Hey JM - nice to see you on the boards...AND...welcome. Been using unRAID myself for several years (12Tb system) and have been very happy with it - but don't ask me how it all works - more mysterious than xbmc by far. JoeL is definitely "Da Man" when help is required and has helped me (and many others) more than words express (Thx Joe). Hope you have fun with it. Flambot (a.k.a Kidkiwi)

Thx - Appreciate the update. I'd certainly like to get my parity speeds up - takes about 18hours. As I don't have that type of MB, guess another option is the go.

Sorry if this is little off topic but, what was the reasoning for swapping out the card? Have you noticed any difference with the new one? Does it plug into the same slot type as the TX4's? Any problem updating? I have two TX4's in my own machine, but would like another option as I couldn't find any locally in NZ.

I also use a 100mbit router and a Gigabit switch. Works great.

Under Linux, yes. Bummer :'(

No and you won't.... ACPI doesn't provide for it. Even for drives with power management that will power up in spin-down mode, it won't work. Does this mean WD drives jumpered to "Not Spin up After Resume from S3" with still spin up - or does it only spin up HDD's connected via ACPI?

This looks great. Does this mean you can access it via a windoze pc via network places?? I guess this would mean the iso can be seen on the network regardless of what device is looking?? Hey JoeL - any chance you could make it so the iso can be switched from "Read Only" to "Read/Write" (hopefully with a check box or password option so it can't be written to default). I tweak a lot of IFO files and this functionality would mean I could write the IFO's directly back to the iso without all the to-and-froing of iso-to-vob-to tweaked vob-back to iso etc.

AGW - Thx for the write-up. Much appreciated.

From memory, I'm going to say 8-10 seconds or so? I definitely remember it being faster than a fresh boot. But again, I'm only running two drives, so not sure how the numbers might change for some of the large arrays that many are running. Of course, after several days of glorious S3 / WOL use - I dug back into this to write-up my FB,BB (For Beginners, By Beginners) Guide to S3 / WOL . . . and I somehow got things all goofed up. So, this weekend I get to wrestle with two things: 1. Server is no longer responding to magic packet. I'm hoping I just made a bad BIOS tweak - I don't remember doing that but I was in the BIOS the other day and it's plausible. 2. When it is sleeping and waking properly, I noticed that the server no longer negotiates a gigabit connection upon wake-up. It's 1000Mb/s off of cold-boot, but if I check it after wake from sleep, it seems to be coming back at 100Mb/s. It's onboard Realtek 8168b NIC, so I suspect a driver issue - which means I probably won't be able to do much about it other than hope a driver update in next unRAID release cleans it up. Or buy a NIC, but I'm inclined not to go that route if I can help it. AGW AGW - Hope you get it sorted out. Thx BQ for your comments - very helpful.

I dunno, you will be the first to test! LOL! Oh no...

I looked up what "Powered up in standby" meant. Cool. This looks promising - as virtually all my drives are WD's (they will all be soon). If this is enabled, will the drives spin up automatically (after returning from a standby state) when they are asked to be a network task?, Or does some software thing have to take place after returning from standby to make the drives ready for use? Thx

Thx muchly for the comments. They should get me started. Pity all the drives spin up, but I guess this is what happens on a Windows machine as well?? If I think about it, I'm not sure I want all the drives to spin up (16 of them) just to save a few watts of power. I'm more worried about the long-term reliability of the HDD's - especially as that's were all the $ lay in the system Being able to shut the server down and start it up from the xbox would be good though. Thx again for the help.