| 01:38:04 | <atphoenix> | I'm attempting to get a project docker image running by following https://archiveteam.org/index.php?title=Running_Archive_Team_Projects_with_Docker |
| 01:38:15 | <atphoenix> | I ran 'docker run -d --name watchtower --restart=unless-stopped -v /var/run/docker.sock:/var/run/docker.sock containrrr/watchtower --label-enable. |
| 01:38:15 | <atphoenix> | ' but ended up with a permission denied. Does this need to be run via sudo? |
| 01:39:40 | <atphoenix> | OS Ubuntu (Lubuntu) 20.04 |
| 02:44:12 | <Craigle> | You can short cut with sudo, or create a docker group. Instructions here: https://docs.docker.com/engine/install/linux-postinstall/ |
| 02:44:42 | <Craigle> | I just run as root on Hetzner machines, since they're ephemeral anyway |
| 02:45:09 | <Craigle> | But on a permanent/real machine, I create the group |
| 03:09:07 | <atphoenix> | thanks, got it working with sudo. Docker on Linux on VM dedicated to AT project. |
| 08:06:55 | | VADemon joins |
| 09:00:22 | | britmob2 quits [Quit: britmob2] |
| 13:50:45 | | britmob2 joins |
| 14:23:08 | | VADemon quits [Client Quit] |
| 14:52:32 | | Arcorann_ quits [Ping timeout: 252 seconds] |
| 15:49:50 | | VADemon joins |
| 17:33:27 | <atphoenix> | is 'docker stop watchtower archiveteam' a clean operation? i.e. does it cleanly wrap up and submit whatever archiving it may currently be doing? |
| 17:35:33 | <Craigle> | Watchtower doesn't need to be stopped cleanly since it doesn |
| 17:35:49 | <Craigle> | since it doesn't have any data to keep/submit |
| 17:36:43 | <Craigle> | For project containers, I usually either manually do a "touch STOP" or docker ps -q | xargs docker stop -t 28800 |
| 17:37:22 | <Craigle> | That command will stop issue a graceful stop to all docker containers, and then forcefully kill them if they are still running at 8 hours |
| 17:37:34 | <Craigle> | The time can obviously be adjusted |
| 17:38:39 | <atphoenix> | Thanks for the details. The command I pasted here was from https://www.archiveteam.org/index.php?title=Running_Archive_Team_Projects_with_Docker |
| 17:41:33 | <Craigle> | Ok, we may want to expand that. It's not wrong, but it won't finish up any running jobs or uploads before killing the container. But that probably isn't the end of the world either |
| 17:41:36 | <atphoenix> | I don't want in-flight work to be lost in case I need to schedule a host restart. |
| 17:51:05 | | themadpro (themadpro) joins |
| 17:58:15 | <@JAA> | We always requeue items at the end anyway, so that shouldn't be an issue. |
| 18:33:04 | | VADemon quits [Ping timeout: 252 seconds] |
| 18:35:21 | <atphoenix> | if a docker container is restarted (the , does it resume where it left off? The basic question is...are there any special considerations to take into account if a docker host needs to be shutdown for a short period for maintenance followed by a restart? |
| 18:35:47 | <atphoenix> | (ignore the 'the ,') |
| 18:39:23 | <Craigle> | The containers are ephemeral, and really can be thought of almost like a template. When you stop the container, consider it gone. When you "restart" a container, you are essentially starting fresh from the template. |
| 18:40:03 | <Craigle> | There are ways to hold containers and such, but it's not worth it for this, and it's really not how docker is intended to run. |
| 20:04:19 | | drd joins |
| 21:46:59 | <tech234a> | docker stop by default gives a 10 second grace period before forcibly stopping the container. The length of the grace period can be controlled using the --time or -t parameter. https://docs.docker.com/engine/reference/commandline/stop/ |
| 21:55:53 | <tech234a> | I updated the tutorial based on the discussions in here, thanks for the feedback! https://archiveteam.org/index.php?title=Running_Archive_Team_Projects_with_Docker |
| 21:56:48 | <Craigle> | Looks good, thanks tech234a !! |
| 21:57:44 | <tech234a> | Should I still keep the message indicating that the page is a draft? |
| 21:59:28 | <tech234a> | I'll leave it for now until a few more people use the tutorial |
| 22:38:44 | <atphoenix> | I have one minor Tracker dashboard.js suggestion ( https://github.com/ArchiveTeam/universal-tracker/blob/588154aa66ce9acc54fd9b80b6ccf8a6ac15c579/public/js/dashboard.js ): Change MiB/u to MiB/item or MiB/i for consistency. (Elsewhere in the WebUI things are referred to as 'items'.) |
| 23:12:27 | | Arcorann (Arcorann) joins |
| 23:12:37 | | Matthww quits [Read error: Connection reset by peer] |
| 23:12:53 | | Matthww joins |
| 23:13:43 | <atphoenix> | One more minor suggestion ( https://github.com/ArchiveTeam/universal-tracker/blob/588154aa66ce9acc54fd9b80b6ccf8a6ac15c579/public/css/dashboard.css ): Make edit to prevent the phrase 'to do' from getting cut off. It is cut off on at least some active projects. Perhaps changing 'margin-left: 450px;' to something slightly larger (460?) will do the trick? |