[15:45] <leumashm> hi guys, i have i query, i know if i have want to allow a ftpuser to read write to /var/www/html i need to do this sudo chown ftpuser:ftpuser /var/www/html but what if i have multiple users ftpusera ,b ,c how do i do it? I see online ppl talking about putting ftpusersa,b,c in groups.
[15:46] <oerheks> normally one makes that user member of www-data:ww-data
[15:46] <oerheks> or group, ftp-users
[15:48] <leumashm> oerheks hmm then how to apply permissions to the /var/www/html ?
[15:49] <oerheks> when they are member of www-data, correct permissions are applied?
[15:49] <leumashm> oerheks any drawbacks for putting users into groups? like after the ftpuserb create the file in /var/www/html he becomes the owner, will it cause problems where ftpuserC cannot delete the file?
[15:50] <oerheks> he does not become owner, the group is.
[15:51] <oerheks> debian page; https://wiki.debian.org/SystemGroups ..
[15:52] <leumashm> oerheks chgrp -R <-groupname> <-dirname->
[15:52] <leumashm> chmod -R g+rw <-dirname->  ?
[18:00] <foo> I just had a droplet crash and noticed one of my services didn't come up but everything else did. 18.04 here. I don't see a cron job restarting services on reboot... and I don't see things in /etc/init.d - so I'm wondering how I got some of these to start
[18:00] <foo> I don't see anything specific in the service
[18:01] <foo> I wonder if I did it with update-rc.d
[18:02] <foo> Can I list the current status of something with update-rc.d ? /me checks man page
[18:03] <oerheks> systemctl status <name>.service
[18:03] <foo> oerheks: thank you, I'm trying to find out if A) a service died trying to start up (sad day) or B) it was never configured to start up
[18:04] <foo> I have two services here, one started up on reboot one didn't
[18:04] <oerheks> with systemd, you would have a unit file?
[18:04] <sdeziel> foo, what do you get from "systemctl is-enabled $service" ?
[18:04] <foo> Is there a way with update-rc.d to show if something should have started up on boot? Unless there's another way this started up
[18:04] <foo> oerheks: yup, I do
[18:05] <foo> sdeziel: I see the service that's running on bootup is enabled, I see disabled for the service that did not start at boot
[18:06] <sdeziel> foo: OK then you can enable it with "systemctl enable $service" and it should start on the next boot/reboot
[18:06] <foo> sdeziel: thank you.
[18:06] <sdeziel> foo: and you can enable and start it in one go: systemctl enable --now $service
[18:06] <sdeziel> yw
[18:47] <RoyK> sdeziel: yw?
[18:47] <sdeziel> RoyK: you're welcome
[19:33] <RoyK> sdeziel: danke
[20:53] <sdeziel> I have a situation where "df -h /" says 15G is used but "du -smc /*" says total is ~7G
[20:55] <sarnold> does lsof show any 7G deleted files in use? :)
[20:55] <sdeziel> I'm fishing for a directory that was potentially deleted but for which a process keeps files open in it, any ideas/pointers on how I could find the process still pinning the files under that dir?
[20:56] <sdeziel> sarnold: I know I'm looking for many files that were under a directory that 'vanished' under us
[20:56] <sdeziel> so I'm not looking for a big file that'd be deleted but a dir... not sure that's even possible
[20:57] <sdeziel> lsof isn't installed and I didn't want to install anything on the FS, fearing side effect but I'll do it
[20:57] <sarnold> no ..
[20:57] <sarnold> fuser?
[20:57] <sarnold> there's also /proc/*/fd/
[20:57] <sarnold> and /proc/*/cwd
[20:58] <sarnold> it's also possible your du command is executing in a different filesystem namespace than other processes
[20:59] <sdeziel> "lsof | grep $dir" => nothing :/
[21:00] <sdeziel> "readlink /proc/*/cwd | grep $dir" => nothing
[21:03] <sarnold> :/
[21:04] <sdeziel>  /proc/[0-9]*/fd/ wasn't better :/
[21:06] <sdeziel> sarnold: thanks for the help!
[21:07] <sarnold> sdeziel: bummer :(
[21:08] <sdeziel> it's OK, it was a stage VM ;)
[21:08] <sdeziel> I would have wanted to understand the discrepancy though
[21:15] <sarnold> sdeziel: yeah, I don't like not knowing *why* things happen