Factorialandha

unfortunately as this is multiple dockers and Ubiquiti dont actually provide a UNMS piece of software, i believe it would be difficult to try to get it into a single traditional docker at least not easily. Would be nice if someone could do it, but ive not managed to find enough documentation or information on how to create dockers for unraid to even try myself.

This would be perfect, however unfortunately UNMS is not one docker, its multiple. I have however managed to get it working using docker-compose via ssh. Its not ideal but my none detailed version is below 1. Build VM 2. Install UNMS using the scripts inside the VM 3. Copy all files created (dockerfile etc in /home/unms) to a folder on your unRaid machine (i used /mnt/user/appdata/unms) 4. Created a user and group within unraid - i used (groupadd -g 1002 unms) for the group and (useradd -u 1002 -g 1002 -d /mnt/user/appdata/unms/ unms) for the user 5. changed ownership on all files in /mnt/user/appdata/unms (just for consistancy) I then i had to modify the docker file, see attached my example docker file the items i changed were related to network (i was already using ports so i decided to give the public dockers there own public ip address my amendments are for the sections (modifications should be pretty clear based on reading through your file) networks: br0: external: true internal: internal: true for this then, everywhere you see "public" change to "br0" under networks, example below, you also need to modify the "volumes" section and replace /home with whatever path you set up. and also the UID (1002 in my case) needs to be updated wherever you see that. fluentd: container_name: unms-fluentd build: context: ./fluentd dockerfile: Dockerfile args: - http_proxy= - HTTP_PROXY= - https_proxy= - HTTPS_PROXY= - no_proxy= - NO_PROXY= # if we specify an image name, docker-compose will try to pull it from repository # see https://github.com/docker/compose/issues/3673 # image: ubnt/unms-fluentd:local restart: always networks: - br0 ports: - "127.0.0.1:24224:24224" volumes: - /mnt/user/appdata/unms/app/conf/fluentd.conf:/etc/fluent/fluent.conf - /mnt/user/appdata/unms/data/logs:/fluentd/log environment: - FLUENTD_UID=1002 the only additional bit would be if you want a static ip (which you probably do) for the unms interface which case you want the below, you can see the networks section is slightly different, in this case i want my ip to be static at 192.168.1.253. the reason for this is that you want your ubi devices to be able to connect to it, in my case i wanted a static ip specifically for the docker because ports 443 and 80 were already in use, and i didnt want to mess around trying to get UNMS working with different ports. nginx: image: ubnt/unms-nginx:0.11.3 container_name: unms-nginx restart: always networks: br0: ipv4_address: 192.168.1.253 volumes: - /mnt/user/appdata/unms/data/cert:/cert - /mnt/user/appdata/unms/data/firmwares:/www/firmwares depends_on: - unms - fluentd logging: driver: fluentd options: tag: "nginx" fluentd-async-connect: "true" environment: - NGINX_UID=1002 - SSL_CERT= - SSL_CERT_KEY= - SSL_CERT_CA= - HTTP_PORT=80 - HTTPS_PORT=443 - WS_PORT= - UNMS_HTTP_PORT=8081 - UNMS_WS_PORT=8082 - PUBLIC_HTTPS_PORT=443 - SECURE_LINK_SECRET=YvQfyu8lwfZjuXMCUiE1tXhf1ggBM7leT7GPs5FXRUvgdWvfwxFlpdwS69mn5AUNJpf9Cl1f9wtzqtDTFoglBdEjRz9cI5ijGBKy feel free to ask any questions and ill try to clarify, this obviously doesnt look great in the interface ( see screenshot) dont even ask why its showing minecraft blocks as the image haha. You will have to get the docker-compose ( i added the below lines to my /boot/config/go file curl -L https://github.com/docker/compose/releases/download/1.14.0-rc2/docker-compose-`uname -s`-`uname -m` > /usr/local/bin/docker-compose chmod +x /usr/local/bin/docker-compose then once you have a modified docker-compose file you can run docker-compose up -d and it should create everything needed (i havent tried it but then you can start and stop the vms via the GUI, however they do rely on one another, so you need to remember which order to start them in, however they shoudl recover if you start them in the wrong order anyway)

I agree with itimpi, personally when i did my initial load, i mounted the drives i was moving data from in the machine physically as unassigned disks and moved the data locally that way it was faster than transferring over the network, it also removed any issues with file transfers and network issues. i also did this without a parity.

perfect, installed and configured, thankyou

For your initial transfer you could move remove your SSD from the cache, That way then its not being slammed with the initial 25TB worth of writes. On the note of the "left over crap from downloading", you could put your downloads / complete / incomplete on a seperate "share" that only ever uses the cache, that way then it doesnt get moved to the array, then have a nightly process that "cleans up" the folders, at the end of the day, you want to clean up the folders regularly anyway, otherwise it will just end up getting full of junk. Im new to unraid also but i have my downloader set up to use a specific share for complete and incomplete downloads and its marked as a cache share only. so it /should/ never move any of that data onto the array.

Genius, I did think about trim but stupidly didnt think to run it manually. Not sure why i just assumed it would be done automatically aslong as the file system and the drive supports it. I will set up a cron job to have this run on a regular basis and monitor to see how well things run.

I have searched the forums and cannot seem to find a similar issue. My unRaid web gui / dockers seem to hang when my cache drive is doing a lot of work. I am running Plex, Couchpotato and SABnzbd it seems when SABnzbd is extracting / files are moving around i get the issue. It also seems to happen when i transfer data, see attached network file copy, it goes for a good chunk of the file at full speed and then severely drops. Im wondering if this is possibly a hardware bottleneck (SATA etc) or if its something else. Any input / ideas appreciated, the machine isnt exactly underspec. andha-nas-diagnostics-20170730-2338.zip