melmurp

Members
  • Posts

    47
  • Joined

  • Last visited

Everything posted by melmurp

  1. Is there a setting for this? I just get the Edit/Remove
  2. I've been using https://github.com/binhex/arch-privoxyvpn as a socks style proxy for a VM for quite some time now as I wanted all traffic from that VM routed through the vpn. I just setup Firefox docker using what you described above... I'll leave it on and see if it ever drops
  3. This came up in a google search so thought I'd post this even if the reply is likely too late for the OP 7.x is compose based but they have a repo with the last version of the all in one 6.x I was using cli but you can make a template using this pretty easily docker run -d --name sf7 -p 8000:8000 -e SEAFILE_SERVER_HOSTNAME=<url_here> -e SEAFILE_ADMIN_EMAIL=<admin_email> -e SEAFILE_ADMIN_PASSWORD=<passwd> -e SEAFILE_SERVER_LETSENCRYPT=false -v <storage_here>:/opt/seafile-data docker.seadrive.org/seafileltd/seafile-pro:latest They use a private docker hub.. instructions below https://customer.seafile.com/downloads/ There's also a community all in one 7.x but I couldn't get it to work correctly for PRO just CE
  4. I was looking over on how to add packages so they persist on boot To test it out first I installed the ecryptsFS packages, mozilla-nss and the kernel module I get the prompts and seems to somewhat go right but then I get errors when it tries to mount. I checked if the kernel module was loaded but doesn't seem to exist so I'm guess it's not as easy as just installing the kernel package Errors I'm seeing mount.ecryptfs: Could not open library handle mount.ecryptfs: Key module [openssl] does not have a subgraph transition node; attempting to build a linear subgraph from its parameter list mount.ecryptfs: Key module [openssl] has empty parameter list Some guidance would be appreciated Rather not install a docker to handle this if I can
  5. app_update will pull down a new version validate runs a checksum like thing against the files to make sure they're not corrupted I don't think conan itself has version checking internally... I don't see anything in the menus but I could be blind
  6. Basically have a boolean env var called auto-update then check if true/false. If false then don't run the below command or parse the mod list.. skip right to starting the game with existing data ${STEAMCMD_DIR}/steamcmd.sh \ +@sSteamCmdForcePlatformType windows \ +login ${USERNAME} ${PASSWRD} \ +force_install_dir ${SERVER_DIR} \ +app_update ${GAME_ID} \ +quit I'm sure other mod heavy games would benefit from it as well... Conan is just is a headache as things update a lot so you might have to reboot multiple times a day so player/server are in sync. Making it configurable lets the server dictate the reboots and the players can use the modlist.txt trick to avoid steams forced updates. Thanks for fixing the parms thing so fast.. appreciate it!
  7. Would it be possible to add auto-update as an option instead of hardcoded? If you have a lot of mods going things go south quickly thanks to workshop auto updating Also wondering why the GAME_PARAMS are ignored for Conan? I added "-nosteamclient -game -server" but if I look at the start-server.sh in the docker nothing is added Honestly not even sure those cmds will help... Conan takes so long to restart so I'm trying ways to possibly speed it up or not require restarts Tks
  8. Yup Here are the calls {"level":"info","msg":"executing youtube-dl command: /usr/local/bin/youtube-dl -j --playlist-end 1 https://www.youtube.com/channel/UCnsQjCVzmEd-YvlQZbo6zCg","time":"2019-12-19T00:37:41Z"} {"level":"info","msg":"DOWNLOAD: {https://www.youtube.com/channel/UCnsQjCVzmEd-YvlQZbo6zCg Video And Audio Uploads from 2NE1 mp4 [] 12-19-2019 00:37:48 Channel /channels/%(uploader)s/video/%(title)s.%(ext)s}","time":"2019-12-19T00:37:48Z"} {"level":"info","msg":"downloading video","time":"2019-12-19T00:37:48Z"} {"level":"info","msg":"{ youtube-dl --ignore-errors bestvideo[ext=mp4] downloads/channels/%(uploader)s/video/%(title)s.%(ext)s https://www.youtube.com/channel/UCnsQjCVzmEd-YvlQZbo6zCg}","time":"2019-12-19T00:37:48Z"} {"level":"info","msg":"executing youtube-dl command: /usr/local/bin/youtube-dl --ignore-errors -f bestvideo[ext=mp4] -o downloads/channels/%(uploader)s/video/%(title)s.%(ext)s Looks like it grabs the last file, I'm assuming to update the json record, and it's not using the -f option then it makes the call to grab all the files using the -f with mp4 set. Noticed that if you click "check" on the gui it ignores the json settings and uses the ones in the drop down
  9. Looks like the gui adds "[ext=any]" which isn't a supported option for youtube-dl If you remove it from the config it'll just stick in "[ext=]" which of course also breaks "Any" should just leave out the -f option as youtube-dl defaults to "-f bestvideo+bestaudio/best" which will try to get the best of both and convert them or grab the best if it's missing conversion tools
  10. That's reasonable and I appreciate the reply... I know most providers tend to use a proper host name and not a direct IP so I can understand why it would be done this way. I just wanted to confirm it's as designed before I poke around Again, thanks for all the work you do with these containers... you're making people's lives much easier!
  11. Could I get some insight into this? I like the idea of having the all in one (privoxy/socks/openvpn) in a single container but defeats the purpose if it removes all the hosts and just hardcodes to the first one.
  12. I'm using a provider that has multiple remote ips (not hostnames) defined in the ovpn file but when the docker spins up it removes them all and inserts the first one at the top of the file so next time it starts only that single entry exist now. Is this intentional? Tks!
  13. Geez, I feel dumb... that's exactly what I was missing. Need to add EXEC to ONBATT in my client config and put an AT ONBATT timer in upssched.conf so it kicks off a timer then executes the mapped command/script Tks!
  14. Since 6.7 anytime I reboot from the webui the system hangs... If I check the physical box it's stuck at waiting for graceful shutdown 90s but never actually times out. If I physically power it off at this point it comes up and of course does a parity check... it sat at that prompt for 30mins before I manually turned it off. I'm going to turn on syslog mirroring and do some reboots... anything I should look for? *EDIT: Ofc with logging on now it works fine... had 3 hangs from gui and one from a ups shutdown but now after 5 gui reboots comes back up fine.
  15. Having an issue with shutdown message.... My nut clients get the batt low message but never the shutdown... I do get a both notifications on my phone though from the main server via pushbullet. Turned on local syslog and will do a controlled shutdown later this week but curious if anyone has seen this before? Almost like the server jumps the gun and kills nut before sending out shutdown.
  16. I did check and seems the majority of mine where Linuxserver.io's bot updating dependence libs on the same day this compromised occurred.. bad timing Curious why dockerhub requires write access to github repo if they're just pulling
  17. You have a lot more faith then me I guess... some authors likely don't even know this happened or have things in a code complete mode so don't check their github daily. Guess I'll go check all the recent ones marked as updated and see what exactly changed to relieve my paranoia.
  18. Could not someone have used the tokens to add themselves to the github repo access, modify some code, and let the auto build do it's thing... then we get the "docker has update" notification and those with auto update just pulled a poisoned copy? For example this was in the wild for a few days and last night I noticed 6 of my dockers had updates pending... worrisome in light of this news.
  19. They might.. anytime I see something like that it means a resource was not loaded or got blocked. If you load up the debug console in the browser (F12 in chrome) then refresh under the console tab it might tell you what is missing