| mybalzitch | boy 24.04 server sure acts weird. I'm installing some stuff via apt and its multiple minute timeouts on things like ipp-usb.service is a disabled or a static unit, not starting it. | 00:05 |
|---|---|---|
| mybalzitch | Jun 13 18:06:09 ubuntu systemd[1]: Reloading finished in 61519 ms. | 00:07 |
| mybalzitch | guess it didn't like the timezone being changed? idk | 00:07 |
| mybalzitch | its acting normally again | 00:08 |
| === chris14_ is now known as chris14 | ||
| mjt0k | a package in main can't recommend a package in universe, right? how about suggesting it? | 04:14 |
| mjt0k | ahasenack: heck, that was.. hackish (/me looking at dbbefcf538bfd3d) :) | 05:37 |
| mjt0k | ahasenack: at this point I think it's better to split them by driver, each driver in its own package, - the way how eg redhat does it.. Or just drop gluster entirely :) | 05:38 |
| mjt0k | ahasenack: if I knew about that stuff, I'd definitely import the changes and used them as a base for further refinements in that area, it's easier to do that in debian than when importing into ubuntu | 05:44 |
| mjt0k | ahasenack: I suggest we merge some things to debian to have some syncronisation point, so to say - for both samba and qemu, - and work from there. I wont make package-rearrangement changes for now. Next I suggest creating separate -gluster packages and maybe -ceph too (dunno what's up with that one in ubuntu) | 06:36 |
| frickler | ceph in noble is in a bad state because they decided to use a development version from git in order not to lag behind later, but since then they're fighting with the build chain being broken | 07:26 |
| mjt0k | ceph is also 64bit-only these days. Ok. | 07:47 |
| === zareem3 is now known as zareem | ||
| ahasenack | mjt0k: a package in main can suggest a package in universe, that's fine | 11:57 |
| === ahasenack_ is now known as ahasenack | ||
| mjt0k | ahasenack: how about Recommends? | 13:21 |
| ahasenack | mjt0k: nope, main cannot recommend universe | 13:21 |
| mjt0k | ok | 13:21 |
| sebboh | Hello. Over at work I've got a Ubuntu Server 24.04. It's connected to our Windows domain.. I am going to set up a cronjob to copy a file from a SMB server. Years ago, I might have used `smbclient`.. What should I use today? I can only access the share using a domain user. (I can probably get away with storing the password on disk if need be, it's an application account..) | 15:58 |
| sebboh | I think there is some `smbmount` or `mount.smb` or something, too? But that may be overkill | 16:00 |
| rfm | sebboh, smbclient still exists, plus mount.cifs. I think there's a way of putting the credentials in a file readable only by root and naming that in the mount options, but I haven't looked at it for a long time | 16:09 |
| sebboh | ok, thank you! | 16:21 |
| rfm | sebboh, if you're still here, it occured to me how I'd probably do it: put the mount in /etc/fstab and include the "x-systemd.automount' option to have systemd mount the file system when it's used, and optionally the x-systemd.idle-timeout to have it unmounted after the cronjob finishes. | 19:55 |
| rfm | sebboh, or you could handcraft a mount unit instead of using fstab and systemd-fstab-generator, but I like having all the mount specifications in one file | 19:56 |
| sdeziel | sebboh: for a `cifs` mount, you can indeed put the credentials in a root-only file and the use `credentials=/etc/samba/user-creds.secrets` or similar in your `/etc/fstab` options | 19:58 |
| rfm | I thought it was something like that, thanks for checking | 19:59 |
| sebboh | ah, thank you both! | 20:42 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!