mason Posted July 16, 2011 Share Posted July 16, 2011 Hi, since I use Unraid i have several problems regarding file operations to the servers. I run 4.7 maxed out with 20 drives. So far it's working well, but i store all my tvshows and media files on the server, so I scan them a lot, make changes, you know the drill. Something that bothers me for quiet some time: If I access the server (SMB connection from Ubuntu and W7) while some other file operation is ongoing the whole thing becomes unresponsive. It takes 10-30 seconds until the operations continue. Am I the only one with this problem? Is there anything I can tweak? My understanding is that unraid is also used for small workgroups, but I really can't believe that since it's locking up constantly with more than one operation to the server. Thanks Quote Link to comment
Falcon Posted July 16, 2011 Share Posted July 16, 2011 I have the same problem with freezing if I have several file jobs at the system. I thought the unraid filesystem is like this. I have not checked if this is only on write operations. It is not always freezing but sometimes. Quote Link to comment
BRiT Posted July 16, 2011 Share Posted July 16, 2011 Please provide details. syslog? drives spun-down? drives idle? drives spun-up? drives pata? drives sata? drives controllers? motherboard? cpu? memory? file operations? Quote Link to comment
Whaler_99 Posted July 16, 2011 Share Posted July 16, 2011 We need more info - what is the hardware you are using for unRaid? What does your network/LAN look like? Post a syslog. Could be any number of things... See this... Shawn Quote Link to comment
mason Posted July 17, 2011 Author Share Posted July 17, 2011 Maybe I was a little vague here, I wasn't after a solution for the problem. I was more about the generel behaviour of the server since I thought, like Falcon, this is the normal behaviour of the whole system since i have this from day one?! Quote Link to comment
abs0lut.zer0 Posted July 17, 2011 Share Posted July 17, 2011 my two cents how i see this on this page.. http://lime-technology.com/technology it quotes.... Ease of installation Easy capacity expansion. Ease of integration with other systems. Easy recovery from disk device failure. Immunity from catastrophic full data loss. Good multiple-stream read performance. Adequate write performance. so i have always thought in the logic of reading and writing to a single drive.... as it all depends on what type of allocation method you use as well... eg. i use high water so therefore if i do multiple copies at the same time they will go to the same drive and it does also hang for me for a while.. like i said just my interpretation of unraid.... Quote Link to comment
Falcon Posted July 17, 2011 Share Posted July 17, 2011 Maybe I was a little vague here, I wasn't after a solution for the problem. I was more about the generel behaviour of the server since I thought, like Falcon, this is the normal behaviour of the whole system since i have this from day one?! Yes I am also after the same answer and don't see this as a problem. I think the reason for the freeze is an combination of sleeping drives and the design with one fixed parity disk. Quote Link to comment
abs0lut.zer0 Posted July 17, 2011 Share Posted July 17, 2011 I think the reason for the freeze is an combination of sleeping drives if it was sleeping drives then this thread should not exist as this is nothing to do with unraid's disk performance (if you slept the disks you should expect the delay) Quote Link to comment
BRiT Posted July 17, 2011 Share Posted July 17, 2011 No, that is not normal operations. (At least not in my experience with my array) I ask for details so we can spot the issue(s) in your situation and make corrections. Without the details you might as well be saying "My shit don't work. Is this normal?" and then having others chime in with "My shit don't work either. It must be normal." which does no one any good. There are certain circumstances that are known that can cause your issue(s), and in most cases they can be corrected. So once again, POST DETAILS! Quote Link to comment
Falcon Posted July 17, 2011 Share Posted July 17, 2011 I will try to reproduce and publish the logs. Quote Link to comment
Johnm Posted July 17, 2011 Share Posted July 17, 2011 sounds like "cache_dirs" might be handy here. Wont know until more info is posted though. Quote Link to comment
Falcon Posted July 26, 2011 Share Posted July 26, 2011 Now I have done some testing and I have also updated to Version 5.Beta 9. and have still total freeze on the system, but I think I have found the reason: I have a motherboard from ASUS P7P55D-E http://www.asus.com/Motherboards/Intel_Socket_1156/P7P55DE/ on this board it is 9 SATA ports and 2 ATA ports. it is separated like this: Intel® P55 Express Chipset built-in 6 xSATA 3.0 Gb/s ports Intel Matrix Storage Technology Support RAID 0,1,5,10 JMicron® JMB363 PATA and SATA controller 1 xUltraDMA 133/100/66 for up to 2 PATA devices 1 xExternal SATA 3Gb/s port 1 xSATA 3Gb/s port (black) and Marvell® SATA 6Gb/s controller: - 2 x SATA 6.0 Gb/s ports (gray) Normaly I have connected the Cache drive to port 6 on the Intel® P55 Express Chipset built-in ports on the motherboard. I have just got an new 3TB WD drive and disconnected the cache drive so I can connect the new drive on that port and run the preclear script on the disk before i put it into the raid. Thats means that the preclear script is running and reading data from the disk on port 6 when I tested this the first time. ================================================================== 1.12 = unRAID server Pre-Clear disk /dev/sdj = cycle 1 of 1, partition start on sector 1 = Disk Pre-Clear-Read completed DONE = Step 1 of 10 - Copying zeros to first 2048k bytes DONE = Step 2 of 10 - Copying zeros to remainder of disk to clear it DONE = Step 3 of 10 - Disk is now cleared from MBR onward. DONE = Step 4 of 10 - Clearing MBR bytes for partition 2,3 & 4 DONE = Step 5 of 10 - Clearing MBR code area DONE = Step 6 of 10 - Setting MBR signature bytes DONE = Step 7 of 10 - Setting partition 1 to precleared state DONE = Step 8 of 10 - Notifying kernel we changed the partitioning DONE = Step 9 of 10 - Creating the /dev/disk/by* entries DONE = Step 10 of 10 - Verifying if the MBR is cleared. DONE = Post-Read in progress: 82% complete. ( 2,467,584,000,000 of 3,000,592,982,016 bytes read ) 79.8 MB/s Disk Temperature: 33C, Elapsed Time: 30:47:45 If I try to write any of the other drives which is connected to the SAME controller but another port 1-5 on (Intel® P55 Express Chipset built-in ports on the motherboard) \\10.10.10.10\disk1 or \disk2 ..... the write is freezing and the write operation is not working. If I try to read from the drives it works as it should. If I try to write to any other drives I have on the other interfaces the write is working as it should. After the preclear was finished, I tried to read a file with 4.5 GB size from one drive on port on the controller and write to another disk on another port on the same controller and the same freeze problem was reproduced. My hardware is: I have a motherboard from ASUS P7P55D-E http://www.asus.com/Motherboards/Intel_Socket_1156/P7P55DE/ All drive active 10 GB Memory Dual Core 2.8 Ghz Drives: SATA 2T WDC_WD20EARS-00MVWB0 Paritet 500G 6G ST3500630AS Disk1 1T 15G WDC_WD10EACS-00C7B0 Disk2 1T 26G WDC_WD10EACS-00C7B0 Disk3 1T 32G WDC_WD10EACS-00ZJB0 Disk4 500G 3G SAMSUNG_HD501LJ Disk5 1T 51G WDC_WD10EACS-00C7B0 Disk6 200G 7G ST3200822AS Disk7 320G 18G ST9320423AS Disk14 ATA 300G 8G Maxtor_6B300R0 Disk8 300G 17G Maxtor_6B300R0 Disk9 320G 34G ST3320620A Disk10 320G 16G ST3320620A Disk11 It is very easy to find when you disconnect the cache drive, if any other with the same controller want to test and see if they have the same freeze. syslog.txt Quote Link to comment
Falcon Posted July 31, 2011 Share Posted July 31, 2011 Bump. No comments on this ? Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.