January 7, 20179 yr When I have things downloading/unpacking on sab the unRAID web GUI and all other dockers except sab become unresponsive or very very VERY slow. If I check the unRAID usage stats it shows the CPU at 10-40% usage and RAM less than half used. If I check netdata usage it shows the CPU at 100% usage and RAM the same as the unRAID GUI which seems to line up more with how the server is acting. From what I understand all of unRAID loads up into the ram and runs from there and unRAID should have to highest priority right? So why does the GUI become so unresponsive? Could it do with the speed of my RAM? Also is the unRAID stats page wrong for CPU usage? Just a note to add I have set sab to only use two of the 4 cores and that doesn't seem to work either. It quite annoying that every time anything downloads my stream freezes and I have to wait. One more thing, did I read somewhere that unRAID is working on setting up settings for nice in the dockers page? Sent from my iPad using Tapatalk
January 7, 20179 yr When I have things downloading/unpacking on sab the unRAID web GUI and all other dockers except sab become unresponsive or very very VERY slow. If I check the unRAID usage stats it shows the CPU at 10-40% usage and RAM less than half used. If I check netdata usage it shows the CPU at 100% usage and RAM the same as the unRAID GUI which seems to line up more with how the server is acting. From what I understand all of unRAID loads up into the ram and runs from there and unRAID should have to highest priority right? So why does the GUI become so unresponsive? Could it do with the speed of my RAM? Also is the unRAID stats page wrong for CPU usage? Just a note to add I have set sab to only use two of the 4 cores and that doesn't seem to work either. It quite annoying that every time anything downloads my stream freezes and I have to wait. One more thing, did I read somewhere that unRAID is working on setting up settings for nice in the dockers page? Sent from my iPad using Tapatalk ` Can it be that you are saturating your network connection ? Can you still easily browse the internet when this is going on ?
January 7, 20179 yr Author When I have things downloading/unpacking on sab the unRAID web GUI and all other dockers except sab become unresponsive or very very VERY slow. If I check the unRAID usage stats it shows the CPU at 10-40% usage and RAM less than half used. If I check netdata usage it shows the CPU at 100% usage and RAM the same as the unRAID GUI which seems to line up more with how the server is acting. From what I understand all of unRAID loads up into the ram and runs from there and unRAID should have to highest priority right? So why does the GUI become so unresponsive? Could it do with the speed of my RAM? Also is the unRAID stats page wrong for CPU usage? Just a note to add I have set sab to only use two of the 4 cores and that doesn't seem to work either. It quite annoying that every time anything downloads my stream freezes and I have to wait. One more thing, did I read somewhere that unRAID is working on setting up settings for nice in the dockers page? Sent from my iPad using Tapatalk ` Can it be that you are saturating your network connection ? Can you still easily browse the internet when this is going on ? Yes I can still use the internet just fine. At most it's using 112 Mb/s over a gigbit connection. Sent from my iPad using Tapatalk
January 7, 20179 yr What does the output of top look like? Perhaps your CPU is spending a lot of time waiting for I/O.
January 7, 20179 yr Author What does the output of top look like? Perhaps your CPU is spending a lot of time waiting for I/O. Do I just run that from the main unRAID command line? Also it's all on a SSD so everything should be pretty speedy. Edit: nvm here it is Sent from my iPad using Tapatalk
January 7, 20179 yr Author 65.5 wa SSD or not, it's spending a lot of time waiting for I/O. That still shouldn't effect the unRAID GUI right? I could see it making Plex slow as it has to load things from the SSD. Sent from my iPad using Tapatalk
January 7, 20179 yr While this won't solve the issue it will help some, look into NzbGet to replace SabNZB.
January 8, 20179 yr Author Your load average is around 12 constantly. How many CPU cores do you have? 4 cores 4 threads. While this won't solve the issue it will help some, look into NzbGet to replace SabNZB. I have done quite a bit of research on this actually and nzbGet isn't really better in performance anymore unless you have a really low powered device. And any performance gain if at all isn't worth it to me to have to switch everything over. P.S. I have tried a couple files with nzbGet and the server did the same thing. Sent from my iPhone using Tapatalk
January 22, 20179 yr Did you ever manage to solve this? I've got a 2-core Xeon with hyper-threading and using a SSD for cache. I've got the same issue when sab starts to unpack(from/to the same SSD). The unraid GUI and all other docker GUIs become unresponsive. Below are the figures from 'top' while the GUI was unresponsive. I also ran 'atop' but forgot to copy the stats. But the CPU normally stays well under 20% during unrar if I remember correct. -------------------------------------------------------------------------------------- top - 19:08:17 up 4 days, 50 min, 1 user, load average: 16.17, 9.68, 4.73 Tasks: 298 total, 3 running, 295 sleeping, 0 stopped, 0 zombie %Cpu(s): 0.3 us, 2.0 sy, 0.8 ni, 0.0 id, 97.0 wa, 0.0 hi, 0.0 si, 0.0 st KiB Mem : 4072456 total, 279136 free, 1147720 used, 2645600 buff/cache KiB Swap: 0 total, 0 free, 0 used. 2091124 avail Mem PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 53892 nobody 30 10 18096 3948 3104 D 6.0 0.1 0:20.64 unrar 54293 root 20 0 0 0 0 R 2.3 0.0 0:00.88 kworker/u128:9 31683 root 20 0 0 0 0 S 1.3 0.0 0:02.13 kworker/u128:11 587 root 20 0 0 0 0 S 1.0 0.0 2:37.12 kswapd0 54292 root 20 0 0 0 0 S 1.0 0.0 0:00.24 kworker/u128:8 35986 nobody 20 0 166284 54080 460 S 0.7 1.3 0:06.95 mongod 1562 root 20 0 9908 2796 2112 S 0.3 0.1 16:43.91 cpuload 16321 root 20 0 0 0 0 S 0.3 0.0 0:02.66 kworker/u128:14 47670 root 20 0 98920 12784 116 S 0.3 0.3 1:03.39 supervisord 50362 root 20 0 0 0 0 S 0.3 0.0 9:52.91 kworker/1:1 55749 root 20 0 24980 3172 2468 R 0.3 0.1 0:00.05 top 1 root 20 0 4372 1536 1436 S 0.0 0.0 0:10.61 init ------------------------------------------------------------------------------------------
January 22, 20179 yr -------------------------------------------------------------------------------------- top - 19:08:17 up 4 days, 50 min, 1 user, load average: 16.17, 9.68, 4.73 Tasks: 298 total, 3 running, 295 sleeping, 0 stopped, 0 zombie %Cpu(s): 0.3 us, 2.0 sy, 0.8 ni, 0.0 id, 97.0 wa, 0.0 hi, 0.0 si, 0.0 st KiB Mem : 4072456 total, 279136 free, 1147720 used, 2645600 buff/cache KiB Swap: 0 total, 0 free, 0 used. 2091124 avail Mem You have the same problem. Your CPU is spending 97% of its time waiting for I/O.
January 22, 20179 yr ahh ok thanks, so basically my SSD is unable to read/write the data quick enough? Trying to figure out how that could be mitigated. Perhaps downloading to a another share on a separate disk without caching turned on and then extract files to a share with cache turned on? Still don't understand why the Unraid GUI becomes unresponsive though
January 22, 20179 yr ahh ok thanks, so basically my SSD is unable to read/write the data quick enough? I don't know, but it doesn't sound very likely, unless it's faulty or badly in need of trimming, perhaps. Trying to figure out how that could be mitigated. Perhaps downloading to a another share on a separate disk without caching turned on and then extract files to a share with cache turned on? Maybe try BRiT's suggestion? I don't use Usenet downloaders myself. Still don't understand why the Unraid GUI becomes unresponsive though Because on average there are 16.17 processes waiting for CPU time. Your 2-core (4-virtual core) CPU is rather overloaded. Hence the sluggish response.
January 22, 20179 yr Thanks for the help I'll try BRiT's suggestion Strange thing is I didn't have any problem with my old server(7yrs old) which had an old Atom CPU and running Debian and 6x2TB disks in RAID6. Sure the unrar took longer time to finish, but didn't notice any interruption on applications running. Will look into if it's possible to limit CPU for a Docker App, otherwise I could try to run this in a VM and limit it to 1 vCPU. Decompression will take longer but perhaps that way it shouldn't slow down the whole system.
January 22, 20179 yr Dockers tend to have their own support pages. It might be worth taking a look at the one relevant to your docker.
January 22, 20179 yr If anyone else has the same problem I found this excellent docker FAQ: http://lime-technology.com/forum/index.php?topic=40937.0 It suggested to add --cpuset-cpus=0,1 to limit a container to certain cores and --memory=4G to limit RAM. These are added under 'extra parameters' in the edit section for the specific container and can be found in advanced view mode. I limited mine to just one core and RAM to 1G and see if that fixes it.
March 4, 20179 yr On 1/22/2017 at 2:19 PM, dannen said: If anyone else has the same problem I found this excellent docker FAQ: http://lime-technology.com/forum/index.php?topic=40937.0 It suggested to add --cpuset-cpus=0,1 to limit a container to certain cores and --memory=4G to limit RAM. These are added under 'extra parameters' in the edit section for the specific container and can be found in advanced view mode. I limited mine to just one core and RAM to 1G and see if that fixes it. Did this fix the issue? I occasionally have this problem too.
Archived
This topic is now archived and is closed to further replies.