thestraycat

OK. So i restarted unraid (38 days uptime) and copied the same file and averaged 58MB/s throughout a 45gb single file transfer (which was what i was expecting.) Any idea why a reboot seemed to speed things up?

Thanks for looking.

Here you go! mediaserver-diagnostics-20171108-2221.zip

Thanks i'll disable it and see if it clears up!

Generating it now. Will upload when it completes.

Agree. Any idea of where to start with this one?

Johnnie.Black - These are mine... https://www.cnet.com/products/hgst-deskstar-7k3000-hds723020bla642-hard-drive-2-tb-sata-600-series/specs/ Quoted at 200mb/s+ internal speeds.. Understand there's going to be penalty's but 8 times? tdallen - CPU = 8% RAM = 29% Copy from the cache -> Disk 4 = same speed Move from disk4 -> disk14 (all spinners) = Averaged at 45MB/s

Hi Johnnie - That's what i thought. the 2tb disks (7200rpm | 6gb/s -> 6gb/s HBA's) Should be capable of 50-75MB's on single large file copies from SSD surely... Any idea where to start on why i might be seeing tragic copy times? The destination disk is a 2TB BLANK, the SSD cache has 600gb free on it . (does it need to be pre-cleared at all, not added additional disks since initial installation of unraid) I'll try the same copy to another disk in the array to see if i see similar disk speeds to out rule a dodgy disk.... Disk's SMART info shows as a pass for what it's worth...

Do you have the 'Turbo Write' option enabled? No i don't. I try and keep the array on as low TDP as much as possible. I didn't really think i'd have to enable turbo mode to get a disk transfer speed of more than 25MB/s on a drive capable of around 140-150MB/s The architecture of unRAID means that each 'write' operation actually involves 4 I/O operations Totally. Seems a bit excessive though on a large file transfer right (1 file at 30gb+ 25MB/s) Especially coming from an SSD Cache. Also FYI, your array is composed of 2TB drives. They're pretty slow - you'd get better speeds with high density 6+ TB drives. Totally agree, however that would involve buying 6 x 6TB Disks (cost) simply for the sake of a few extra MB/s for disk to disk copies.... Mine were all free (with a handful of spares ready for disk failure!) and should be more than capable of archiving data. What sort of disk to disk speeds could i expect going from SSD -> 6tb disk without turbo mode? I suppose i'm trying to evaluate whether my config has an issue and whether the copy i detailed should be sitting more around 50/60/70MB/s or whether a copy of a 30GB single file from SSD -> disk is perfectly acceptable to be 25MB/s it just dosnt seem right to me. What speeds do you guys see doing these sort of copies?

Hi guys, Any idea why my native disk to disk copies would be painfully slow? I've tested using MC via SSH and windows SMB copy on a 30gb+ single file copies and i burst around 200MB/s and then continue to fall until i reach 25/30MB/s I was really expecting to be able to hold 70-90MB/s disk to disk across the HBA's and MB! Nothing else happening accross VM's/Docker at the time of the copy. CPU <10% as well. 1tb Samsung Evo (CACHE DRIVE) [Plugged into 6GB/s port on MB] -> Hitachi 2TB (Disk14) [Plugged into 6GB/s HBA] My unraid setup runs dual parity on the following hardware spec: Supermicro X9SCM-F (E3-1240) 3 x Dell H200 crossflashed to IT mode. (using 2 8087 cables per card to midplane) 16 x 2tb disks + 1tb samsung evo cache Anyone experienced anything similar?

Thanks Squid. Appreciate the speedy feedback! I believe that may have been an internet drop out my end. My GF was at home that day and restarted switch and modem. Well spottedd! I have been having a few wifi/internet lock ups over the last few weeks. But noticed the file buffering issue mentioned in the original post yesterday on the 30th of Sept. Nothing on the 30th look suspicious in the same vein?

since posting: I've deleted all unused redundant appdata Noticed that disk: Hitachi_HDS723020BLA642_MN1240F33DH6XD-20171001-2155.txt does seem to have errored early in in the smart report but this looks historic and previous to me using it for unraid. But can anyone see anything else??

Hi guys, So for the last few weeks ive been getting call traces discovered by the fix common problems plugin. I'm running version 6.3.5 of unraid and a handful of dockers and plugins below: I don't have much to go on other than occasional buffering on my media centers (NUC's running LibreElec) It dosnt matter if there running a 400mb file or a 20gb file from Unraid the media normally buffer once or twice during that time. LibreElec is also giving me cache full errors which seems to point to network traffic? Everything is cabled over gigabit so i shouldn't really have these issues. Could a duff disk be the culprit? Has anyone experienced any funny issues running any of the below dockers or plugins? I'm at a loss. my system utilization is low and the issues show themselves at any point in the day when running media from unraid. Any help would be legendary. I've just replaced all 4 sticks of my ram (4 x 8gb DDR3) through kingston as i was getting a few ECC errors from 1 stick in the IPMI log. Since i replace the RAM i've had no issues. I didn't use to get these issues months back. only noticed it when i moved to unraid 6.1/6.2 but that may be coincidence. Dockers: Apache (Linuxserver/apache:latest) cadvisor (google/cadvisor:latest) cops (Linuxserver/cops:latest) crashplan (gfjardim/crashplan:latest) deluge ((Linuxserver/deluge:latest) dokuwiki (sparklyballs/dokuwiki:latest) glances (nicolargo.glances:latest) hyrda (nicolargo/glances:latest) kursader (sparklyballs/krusader:latest) muximux (Linuxserver/muximux:latest) netdata (titpetric/netdata:latest) ombi (Linuxserver/ombi:latest) openvpn-as (Linuxserver/openvpn-as:latest) PlexMediaServer (Limetech/plex:latest) plexpy (Linuxserver/plexpy:latest) radarr (Linuxserver/radarr:latest) RDP-Calibre (aptalca/docker-rdp-calibre:latest) sabnzbd (Linuxserver/sabnzbd:latest) sonarr (Linuxserver/sonarr:latest) Plugins: Advanced Buttons CA Auto Turbo Write Mode CA auto Update Applications CA Backup / Restore app CA Cleanup Appdata CA Resource Monitor Command Line Tool Community Applications Diskspeed Test Dynamix Active Streams Dynamix Local Master Dynamix SSD Trim Dynamix System Info Dynamix System Temp Dynamix System Statistics Dynamix web gui File Activity Fix Common Problems IPMI Support Nerd Tools Open Files Tips and Tweaks Unassigned Devices unBALANCE unraid Server OS Wake on LAN support DIAGNOSTIC ATTACHED. mediaserver-diagnostics-20171001-2155.zip

Hi guys - A quick question about bulk book import... I'm currently suffering from having my docker.img file totally fill up on large ebook imports... I run a 50gb Docker ISO image... And when i point a large zip file of comics etc to Calibre it fills it up very quickly... Would i be right in assuming that /tmp for calibre needs to be mapped outside of the image? It seems cailbre ebook import grinds to a half when the docker image is totally full (not to mention it effects other running containers) Havnt been able to confirm which temp folder calibre uses to do its mass imports but was assuming /tmp from within the container. I know calibre also has variables that can define this path https://manual.calibre-ebook.com/customize.html but not sure whether they are available to be used when running calibre as a container and whether they need to be built/passed through the unraid template to be used? Any help would be awesome...

Amazing! - Really looking forward to getting this up...

I get the same error... Anyone got a fix for where to set the hostname?

Firstly sorry for the newb question... But Interested in installing the pack purely for iperf but was a little worried about it modifying/removing any dependencies that unraid may rely on although im sure they install to a different place? Not quite sure how 'plugins' install in terms of isolation from the OS but assume they dont have the isolation of containers or VM's at all? Just looking for peace of mind really that there safe to install and remove at a later date without doing damage to the system in anyway?

I'm assuming that radarr's dronefactory wont process folders/files that havn't been created within Radarr. As in radarr wont auto process and auto import content from the dronefactory and make an entry for the movie unless it was prestaged in radarr?

Hi Squid - Totally agree and that side is fine. Importing and processing of media works like a dream between radarr, sonarr, sabnzbd and deluge that's not the problem for me. It's using the dronefactory feature as auto-processing watch folder for downloads that happen outside of radarr/sonarr that is the issue.

Pretty sure i've tried it but i'll give it a go now! I take it you mean change it on radarr, sonarr and sab? UPDATE: Just turned off radarr. Remapped the /downloads -> /mnt/cache/appdata/downloads/ Havn't done sabnzbd or sonarr yet as they shouldn't have anything to do with this test. and still nothing. Clicking on the drone factory button under "wanted" did take a minute or so to finish but nothing was processed. I have around 160 folders (each a movie in the drone factory at the moment for what it's worth, ive also tried moving them all out and keeping 1 film in the folder and restarting) Nothing. Can anyone post there container mappings, radarr/sonarr drone factory path and potentially there remote mappings for me to try emulating? UPDATE: Renamed the old radar drone factory -> radardrone2) created a new folder called radardrone. Copied 1 movie with a reasonably nice name into this folder for the radardrone to post process. Nothing. And no mention in the logs about it from forcing the radardrone factory to scan

Hi guys, i have a weird issue with Radarr/Sonarr and the drone factory not picking up folder for processing (not sure if it's even still ment to be working as there was talk of depreciating it?) My paths are set as below in unraid: Sonarr (Unraid config) /downloads --> /mnt/user/appdata/downloads/sonardrone/ Radarr (Unraid config) /downloads --> /mnt/user/appdata/downloads/radardrone/ And within Radarr/Sonarr GUI, under dronefactory, there set as follows: Sonarr Drone Factory (Sonarr GUI) /downloads/sonardrone/ Radarr Drone Factory (Radarr GUI) /downloads/radardrone/ Everything's on the same server (192.168.1.125) and i'm fresh out of ideas! nothing in the logs corresponding to anything useable! I threw some attachments in to show, i can browse to the folders from within Radar and Sonar and can see the folders of films waiting to be processed but for some reason nothing happens when i manually trigger the dronefactory? Does the config look right to you? I'm starting to think the feature has been quietly depreciated... Can you see anything strange in the screenshots below?

Hi guys, I'm having some strange permission issues with Sonarr. When sonarr creates files in the shares i have created. it creates the files as: Owner Name: nobody Group Name: users sonarr is set with defaults as far as i can tell: Set permissions = Yes File CHMOD mask: 0644 Folder chmod mask: 0755 however from a windows machine i cant edit the files it creates: "you require permission from Unix user\nobody to make changes to this file." Any ideas? I've checked the file permissions from unraid for files being created by Couchpotato and the files show up the same as they do for files created by sonarr (ownername = nobody | Group Name = users) And files editted from windows created by couchpotato edit fine. From windows the couchpotato files look like this: Security tab on a file shows: Everyone - Read & Execute, Read, Write, Special permissions greyed nobody - Read & Execute, Read, Write, Special permissions greyed users - Read & Execute, Read, Write, Special permissions greyed however in sonarr they look like this: Everyone - Read nobody - Read, Write users - Read The shares look identical from within Unraid any ideas why NTFS/ACL's arn't being set properly? i've tried setting sonarr as Set permissions = Yes File CHMOD mask: 0777 Folder chmod mask: 0777 and Set permissions = No File CHMOD mask: 0777 Folder chmod mask: 0777 Am i ment to be specifying the following values within sonarr with "99" & "100" or anything? chown user chown group

Thanks Johnnie

Great stuff. It's just a matter of reassigning the new slot to the cache drive when it comes back up?

Right - Thanks for the confirmation - What are the consequences of powering down -> swapping over the SSD to mainboard SATA0 port -> turning on Will i need to play around with anything prior?