IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 23 October 2018 (all times are UTC)
| 00:04 | GodFather has left IRC (GodFather!~rcc@2600:1009:b029:237a:e156:e8e4:188c:661b, Ping timeout: 264 seconds) | |
| 00:08 | lucascastro has left IRC (lucascastro!~lucascast@177-185-139-186.isotelco.net.br, Remote host closed the connection) | |
| 00:41 | adrianor1 has joined IRC (adrianor1!~adrianorg@187.113.246.204) | |
| 00:45 | adrianorg has left IRC (adrianorg!~adrianorg@177.18.50.89, Ping timeout: 264 seconds) | |
| 03:02 | GodFather has joined IRC (GodFather!~rcc@174-081-217-069.dhcp.chtrptr.net) | |
| 05:40 | kjackal has joined IRC (kjackal!~quassel@2a02:587:3101:400:25fc:2b81:5716:a939) | |
| 06:00 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Ping timeout: 252 seconds) | |
| 06:01 | ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz) | |
| 06:02 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 06:11 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Ping timeout: 264 seconds) | |
| 06:24 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 06:35 | kjackal has left IRC (kjackal!~quassel@2a02:587:3101:400:25fc:2b81:5716:a939, Ping timeout: 252 seconds) | |
| 07:29 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Read error: No route to host) | |
| 07:30 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 07:47 | jgee has left IRC (jgee!~jgee@190.159.118.121, Ping timeout: 252 seconds) | |
| 07:55 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Read error: Connection reset by peer) | |
| 07:56 | alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg) | |
| 09:08 | kjackal has joined IRC (kjackal!~quassel@2a02:587:3101:400:8c1b:8441:9019:ed7a) | |
| 09:22 | Da-Geek has joined IRC (Da-Geek!~Da-Geek@135.196.42.4) | |
| 09:34 | Da-Geek has left IRC (Da-Geek!~Da-Geek@135.196.42.4, Quit: Leaving) | |
| 10:17 | adrianorg has joined IRC (adrianorg!~adrianorg@186.213.153.215) | |
| 10:20 | adrianor1 has left IRC (adrianor1!~adrianorg@187.113.246.204, Ping timeout: 252 seconds) | |
| 10:23 | <enaut[m]> vagrantc: would you be willing to look over the things I did to ltsp-manager? Any "do's" I missed any "don'ts" I did?
| |
| 10:24 | See this branch: https://code.launchpad.net/~franz-die/ltsp-manager/+git/ltsp-manager/+ref/testing-meson
| |
| 10:27 | alkisg, vagrantc: I'm thinking about moving ltsp-manager forward to the account manager state. I looked around on how to better implement the root tasks of adding and removing users/groups
| |
| 10:28 | I found that the dbus interface org.freedesktop.Accounts offers account creation and deletion but not group management.
| |
| 10:29 | It might be planned to do all this with the sssd service but thos plans are not yet in code.
| |
| 10:31 | So I think it would be best to write an own dbus service (that also implements the file-sharing) and uses the existing dbus interfaces where possible.
| |
| 10:32 | That said this new account-manager will need a new name :) - do you have any suggestions for this?
| |
| 10:37 | Oh and if any improvements to my suggestions could be made I'll be happy to consider them.
| |
| 11:23 | <alkisg> enaut[m]: yeah I think we've looked into accountsservice in the past, and found it was inadequate then too, that's why we resorted to adduser/deluser etc
| |
| 11:23 | Writing your own dbus service just for your own app... i'm not sure if that has any benefits
| |
| 11:38 | spaced0ut has left IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut, Ping timeout: 252 seconds) | |
| 11:42 | lucascastro has joined IRC (lucascastro!~lucascast@177-185-139-186.isotelco.net.br) | |
| 12:01 | <enaut[m]> Well that dbus service would run as root while the app can be run as normal user... requiring a password when needed. - or did I missunderstand something there?
| |
| 12:02 | Faith has joined IRC (Faith!~Paty_@unaffiliated/faith) | |
| 12:09 | <alkisg> enaut[m]: ah sure, if you're talking policykit and running gui apps without root etc etc, you're right
| |
| 12:10 | A dbus service can also function as a semaphore between multiple guis, i.e. that they wouldn't be able to add the same user at the same time
| |
| 12:12 | <enaut[m]> yep thats exactly the thing... I doubt that I get it to the general gnome user service though
| |
| 12:13 | <alkisg> As for name... how about gum, a gtk user manager :P And it's easy to find an icon for that :P
| |
| 12:17 | <enaut[m]> sounds better than everything I have come up with
| |
| 12:35 | GodFather has left IRC (GodFather!~rcc@174-081-217-069.dhcp.chtrptr.net, Ping timeout: 252 seconds) | |
| 14:05 | spaced0ut has joined IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut) | |
| 15:31 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 16:07 | GodFather has joined IRC (GodFather!~rcc@2600:1007:b02a:abdd:e156:e8e4:188c:661b) | |
| 16:20 | spaced0ut has left IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut, Ping timeout: 264 seconds) | |
| 16:20 | spaced0ut has joined IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut) | |
| 16:28 | GodFather has left IRC (GodFather!~rcc@2600:1007:b02a:abdd:e156:e8e4:188c:661b, Ping timeout: 250 seconds) | |
| 16:43 | bwicksall has left IRC (bwicksall!~bwicksall@fw.pls-net.org, Ping timeout: 252 seconds) | |
| 16:46 | GodFather has joined IRC (GodFather!~rcc@2600:1007:b02a:abdd:e156:e8e4:188c:661b) | |
| 16:51 | GodFather has left IRC (GodFather!~rcc@2600:1007:b02a:abdd:e156:e8e4:188c:661b, Ping timeout: 264 seconds) | |
| 16:57 | <mwalters> yaum
| |
| 17:03 | lucascastro has left IRC (lucascastro!~lucascast@177-185-139-186.isotelco.net.br, Remote host closed the connection) | |
| 17:03 | GodFather has joined IRC (GodFather!~rcc@198.111.167.146) | |
| 17:27 | adrianor1 has joined IRC (adrianor1!~adrianorg@177.134.56.11) | |
| 17:30 | adrianorg has left IRC (adrianorg!~adrianorg@186.213.153.215, Ping timeout: 244 seconds) | |
| 17:32 | lucascastro has joined IRC (lucascastro!~lucascast@170.78.53.20) | |
| 17:45 | adrianor1 has left IRC (adrianor1!~adrianorg@177.134.56.11, Ping timeout: 240 seconds) | |
| 17:46 | <spaced0ut> when im done with this project i'm going to contribute to current documentation if you guys want. an examples section would be incredibly useful. i cant imagine more than a few paths someone would want to start out on
| |
| 17:47 | <alkisg> You'll be surprised how many say that and when they're finished we don't hear back from them :D
| |
| 17:47 | <spaced0ut> probably cause they never figure out how to finish
| |
| 17:47 | * mwalters whistles | |
| 17:48 | <mwalters> Not sure if anyone cares, but ltsp seems to play nice freeipa
| |
| 17:48 | and vice versa
| |
| 17:48 | I was able to log in with a freeipa user on a fatclient
| |
| 17:48 | screen locking worked
| |
| 17:49 | it's definitely a bit of a bear to set up (at least on ubuntu...)
| |
| 17:49 | <spaced0ut> that sounds very useful for some use cases. i'm still trying to figure out the most simple one. a single autologged in user where the home directory is set in the chroot but changes made by clients are not saved.
| |
| 17:50 | i can't find any examples on COW and the nbd-server man page is flat out broken
| |
| 17:51 | <mwalters> You jsut want the desktop reset everytime?
| |
| 17:51 | <spaced0ut> yes
| |
| 17:51 | <mwalters> do you want temporary storage in the home dir?
| |
| 17:51 | <spaced0ut> i dont care. as long as its not permanent storage
| |
| 17:51 | <mwalters> so they don't need to be able to save stuff
| |
| 17:51 | <spaced0ut> nope
| |
| 17:51 | <mwalters> make the home dir read only?
| |
| 17:51 | <spaced0ut> in the permissions on the server?
| |
| 17:51 | <mwalters> yup
| |
| 17:52 | <spaced0ut> causes issues with chrome and caja
| |
| 17:52 | <mwalters> oic
| |
| 17:52 | <spaced0ut> does that make sense?
| |
| 17:52 | <mwalters> it does
| |
| 17:52 | I bet chrome yells about the profile being locked or some silliness
| |
| 17:52 | <spaced0ut> i'm sure there would be more issues than that. those were just two things screaming in my face as soon as i booted
| |
| 17:52 | <mwalters> sure
| |
| 17:53 | that was more a random thought, than a firm recommendation ;)
| |
| 17:54 | <spaced0ut> so i modified the file that checks for LOCAL_APPS and if its in the lts.conf it uses sshfs to mount the home directory on the server. i accidentally messed it up the one time and found that an invalid command there gives me almost what im looking for haha
| |
| 17:55 | <alkisg> (08:47:28 μμ) spaced0ut: probably cause they never figure out how to finish ==> no no i'm talking about persons that get help and finish properly
| |
| 17:55 | Over the years I think there were more than 100 of them...
| |
| 17:55 | <mwalters> The only thing I have to contribute is bad advice and maybe an occasional meme
| |
| 17:56 | but I'll chill in the channel at least ;)
| |
| 17:57 | In any case, I think we've decided on ditching our business office windows desktops. Gonna set up our windows server as an RDP server and replace the windows desktops with LTSP boxes with remmina
| |
| 17:57 | adrianorg has joined IRC (adrianorg!~adrianorg@179.179.74.252) | |
| 17:58 | <mwalters> I just can't see an argument for maintaining these 4 windows boxes. Even with WSUS, updates are a nightmare. Previous IT manager didn't update them since Jan 2017... The moment I tried to get them up to date, they started update looping. I'm done with you, MS!
| |
| 17:59 | That'll bring our ltsp client count up to 30 at this office
| |
| 17:59 | spaced0ut_ has joined IRC (spaced0ut_!~spaced0ut@unaffiliated/spaced0ut) | |
| 18:00 | <mwalters> 48 total across all offices I think
| |
| 18:00 | <spaced0ut_> but after taking a step back and looking at what i need... i might just need to pxe boot an image. ltsp makes getting most of the way there extremely easy
| |
| 18:01 | spaced0ut has left IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut, Ping timeout: 268 seconds) | |
| 18:01 | spaced0ut_ is now known as spaced0ut | |
| 18:02 | kjackal has left IRC (kjackal!~quassel@2a02:587:3101:400:8c1b:8441:9019:ed7a, Ping timeout: 260 seconds) | |
| 18:04 | <mwalters> oh, freeipa also solves password resets... has a web interface that regular users can log into to reset their own password
| |
| 18:04 | :%s/reset/change
| |
| 18:05 | hell, I think you can plug in your ssh keys and junk too
| |
| 18:05 | That's p cool
| |
| 18:06 | ...I think they can also delete themselves, out of the box... D:
| |
| 18:06 | ...no, it just lets you click it... lol
| |
| 18:13 | <alkisg> (08:48:36 μμ) mwalters: I was able to log in with a freeipa user on a fatclient (08:48:42 μμ) mwalters: screen locking worked ==> that's because of ldm/ssh, not of freeipa
| |
| 18:13 | The client authenticates via ssh to the server, and the server uses freeipa
| |
| 18:13 | GodFather has left IRC (GodFather!~rcc@198.111.167.146, Ping timeout: 240 seconds) | |
| 18:14 | <mwalters> yeah, I wasn't sure of how that all worked, given the copy of the shadow file or whatever for fat clients
| |
| 18:14 | In any case... so far this seems like a good solution for auth across multiple servers
| |
| 18:15 | It's probably easier to set up if you use fedora or centos as the original devs did ;)
| |
| 18:15 | <alkisg> mwalters: dont you have a master server there? One of the four?
| |
| 18:15 | <mwalters> They're all separated by wan
| |
| 18:15 | they're all independent right now
| |
| 18:15 | <alkisg> So when you add a user to one server, how do the other servers get the user?
| |
| 18:15 | <mwalters> and we just sync the home folders
| |
| 18:16 | I ssh to those ones and add it :D
| |
| 18:16 | <alkisg> Haha, why do you need freeipa then?
| |
| 18:16 | <mwalters> because I want that to not be how it works
| |
| 18:16 | <alkisg> OK, I was asking about the desired result
| |
| 18:17 | I.e. that one of them will need to be the "domain controller" or however else it's called
| |
| 18:17 | <mwalters> Ideally, I just want to manage users in one location. We have some pretty strict termination procedures... doing things 4 times over is just asking to screw it up
| |
| 18:17 | <alkisg> And the others would autosync with that, and use caching
| |
| 18:17 | <mwalters> oh, right now I'm planning on just using a single DC/Domain primary... whatever FreeIPA's terminology is
| |
| 18:17 | <alkisg> Because if you omit the sync/caching, you already can tell ltsp to authenticate against any server you want, even remote ones
| |
| 18:17 | <mwalters> orly
| |
| 18:18 | WAN failure here is pretty uncommon, so the credential caching isn't a huge thing for us
| |
| 18:18 | <alkisg> Sure, e.g. you can have "local" nfs home dirs, and authenticate against the master server
| |
| 18:18 | The authentication is a few kb only
| |
| 18:18 | So it's not a matter of speed or bandwidth
| |
| 18:18 | It's just an ssh
| |
| 18:18 | <mwalters> yeah, we have big fat pipes... 100/100 here, 1gb/1gb at two other remotes... and like 250/10 at the other
| |
| 18:19 | a couple kb is nothing ;)
| |
| 18:19 | <alkisg> Eh, then you don't need freeipa, just plain ltsp
| |
| 18:19 | <mwalters> this is just pam over ssh or something?
| |
| 18:19 | <vagrantc> spaced0ut: i think you could probably just create a hook that creates the autologin homedir before the localapps stuff runs
| |
| 18:19 | <alkisg> !nfs
| |
| 18:19 | <ltsp> nfs: to enable NFS home directories for localapps and fat clients, install nfs-kernel-server on your server, nfs-common on your client (don't forget ltsp-update-image), and put this in lts.conf: FSTAB_1="server:/home /home nfs defaults,nolock 0 0"
| |
| 18:19 | <vagrantc> spaced0ut: then it wouldn't mount the homedir and just use the ram on the client
| |
| 18:19 | <alkisg> This gives you "local" nfs homes
| |
| 18:19 | <vagrantc> spaced0ut: and if you wanted it to be clean between sessions, it could delete it first
| |
| 18:19 | <alkisg> LDM_SERVER=master-server
| |
| 18:20 | This gives you centralized authentication. Done.
| |
| 18:20 | <mwalters> interesting
| |
| 18:20 | and they'd have the same UIDs on the secondary servers?
| |
| 18:20 | (that's another annoyance of the current set up)
| |
| 18:20 | <alkisg> The secondary servers won't care about uuids
| |
| 18:20 | It's just NFS, you can use it without name mapping
| |
| 18:20 | The ltsp clients get their uids from the server, so it's fine
| |
| 18:21 | <vagrantc> as long as all the servers get the same uid/gid ...
| |
| 18:21 | <mwalters> interesting
| |
| 18:21 | yeah, this would be great... freeIPA is a total beast... that we don't *really* need, but managing auth across 4 servers makes me dizzy
| |
| 18:22 | <alkisg> Password setting can be done with remoteapps and gnome-about-me or mate-about-me etc
| |
| 18:22 | <mwalters> So /home still exists across all 4 servers, but is mounted via NFS vs however it's mounted out of the box... and the LDM_SERVER handles authentiacation and rights to the home dirs?
| |
| 18:23 | <vagrantc> is alkisg really suggesting mounting homedirs over the internet over NFS?
| |
| 18:23 | <alkisg> No, they would be local to the locations
| |
| 18:23 | 4 locations, 4 nfs servers
| |
| 18:23 | <vagrantc> got it
| |
| 18:23 | and the 4 servers are kept in sync some other way?
| |
| 18:23 | <mwalters> yes
| |
| 18:24 | think we're using ldsync right now
| |
| 18:24 | * alkisg has helped several people with similar setups... | |
| 18:24 | <mwalters> which runs way faster than I can run adduser on 4 servers... which causes weird permissions results ;)
| |
| 18:25 | <vagrantc> yeah, keeping accounts synced across machines needs real tools
| |
| 18:25 | kjackal has joined IRC (kjackal!~quassel@2a02:587:3101:400:8c1b:8441:9019:ed7a) | |
| 18:25 | <alkisg> Well if the servers are mirrors (starting from the same cloned installation), you could even use cp passwd, as shadow wouldn't matter, only the master server shadow would be used
| |
| 18:26 | (i.e. if you want to be able to match uid/names even locally)
| |
| 18:26 | <mwalters> They're not, as they exist right now... but yeah, I'm cloning a VM for the 18.04 update
| |
| 18:27 | so it'll be a "from scratch" set up
| |
| 18:27 | <alkisg> But the idea would be that the users wouldn't ever login to the secondary servers; no user accounts would be needed there; just home dirs
| |
| 18:27 | <vagrantc> LDM always logs into the same server, but the homedir may be on another server?
| |
| 18:27 | <mwalters> and there's no possibility of those users having access to another user's homedir?
| |
| 18:27 | <alkisg> Yes, if /home/username is already mounted, it's not remounted via ldm
| |
| 18:28 | No, uuids still prohibit them
| |
| 18:28 | <mwalters> xlnt
| |
| 18:28 | <alkisg> *uids
| |
| 18:28 | <mwalters> Nice.
| |
| 18:28 | <alkisg> Meh why do i keep typing uuids :D Too much disk partitioning lately :D
| |
| 18:28 | <vagrantc> simply having the homedir present can prevent the mount, or does it need to be a distinct mountpoint?
| |
| 18:28 | <mwalters> This is a huge help. I really wasn't looking forward to using freeipa ;)
| |
| 18:29 | <alkisg> If the homedir is not in the NBD cow, it's considered to be already mounted, and thus skipped
| |
| 18:29 | stat /home/username vs stat /
| |
| 18:29 | I specifically put the test there for nfs home, local home etc (i.e. even local disks on the clients themselves)
| |
| 18:31 | <vagrantc> ah, that's why in a kiosk you would need to mount something there?
| |
| 18:32 | just having the directory present won't work
| |
| 18:32 | guess you could just mount a straight tmpfs
| |
| 18:33 | i definitely remember setting up a kiosk-like thing at some point, but i didn't use LDM at all
| |
| 18:33 | just lightdm and autologin and created the homedir on boot
| |
| 18:33 | <alkisg> Right, I've done it with a tmpfs in the past using an INIT_COMMAND in lts.conf
| |
| 18:34 | But I prefer a second nbd cow partition as it allows preinitialized content and caches on the server
| |
| 18:34 | (well, unless security is much of an issue)
| |
| 18:37 | <vagrantc> right
| |
| 18:37 | for whatever pre-initialized content you have...
| |
| 18:37 | <spaced0ut> you have an example config of a second nbd cow partition?
| |
| 18:37 | <vagrantc> with no pre-initialized content, may as well use tmpfs
| |
| 18:38 | <alkisg> spaced0ut: you need to use mksquashfs to create one out of the template /home/username
| |
| 18:39 | <vagrantc> could also use a cow mount from another directory in the LTSP image ... but much slower to make updates to the homedir portion
| |
| 18:39 | <alkisg> server-side caching also helps if temp users write a lot to /home/username, e.g. an nbd cow can cache gigabytes, while an nbd swap not
| |
| 18:39 | <spaced0ut> the nbd-server man pages dont even work. there's an examples section that every command responds telling you thats a legacy command and it doesnt work
| |
| 18:40 | so i'm looking for "an example config"
| |
| 18:40 | <alkisg> It's the same as the ones in /etc/nbd-server/conf.d
| |
| 18:40 | except for the cow line
| |
| 18:40 | <spaced0ut> so exportname=/home/user
| |
| 18:40 | <alkisg> Ah, the swap conf there would be extremely similar
| |
| 18:40 | <spaced0ut> how do the clients know to mount that and where?
| |
| 18:40 | <alkisg> With an nbd-client command that you'd put to ldm/rc.d
| |
| 18:41 | So that it's remounted on each login
| |
| 18:41 | (i.e. cleared, caches dropped etc)
| |
| 18:42 | <spaced0ut> update an existing one and add that or create a totally new config there?
| |
| 18:42 | <alkisg> new config
| |
| 18:43 | <spaced0ut> im trying to grep through to see the existing one and cant find it
| |
| 18:43 | this would be something that would go great in the docs. i hardly have a clue how that works and there's little other reason to even know about thsi dir
| |
| 18:44 | <alkisg> We don't have control over the nbd docs...
| |
| 18:46 | <spaced0ut> of course not i understand that but youre tied in so tightly that an example would be insanely helpful
| |
| 18:46 | you have to dig so deep to even find that this directory is where the mounts are configured
| |
| 18:47 | maybe this is over my head and this is common knowledge to others...
| |
| 18:48 | <alkisg> man nbd-server => /etc/nbd-server/config; cat /etc/nbd-server/config => /etc/nbd-server/conf.d
| |
| 18:48 | It indeed needs two steps to get there...
| |
| 18:49 | You can file bug reports against the nbd package
| |
| 18:51 | <spaced0ut> i dont mean to be rude here but those steps you provided dont even come close to explaining how the ltsp image uses the /usr/share/ldm/rc.d/configs to mount the exported nbd-server directories
| |
| 18:53 | <alkisg> There's no configuration in the dir you said; just code
| |
| 18:53 | <spaced0ut> especially when your home directory is already mounted via sshfs from the configs in X01-localapps
| |
| 18:53 | <alkisg> If you mean "documentation on the ltsp code", yeah that's rare :)
| |
| 18:54 | X01-localapps has code that avoids the sshfs mount when the home dir is already there
| |
| 18:55 | <spaced0ut> i see
| |
| 18:56 | <vagrantc> would be pretty simple to implement an option that that doesn't attempt to mount anything...
| |
| 18:56 | <spaced0ut> do you have an example of something in /usr/share/ldm/rc.d/ that mounts another exported directory besides the built image? ive heard it mentioned in here a few times over the last 3 days and cant find a single example when looking it up
| |
| 18:57 | <vagrantc> just write a simple script
| |
| 18:57 | #!/bin/sh
| |
| 18:57 | mount -t tmpfs tmpfs /home
| |
| 18:57 | #end of file
| |
| 18:58 | <spaced0ut> hmm that easy huh
| |
| 18:58 | <vagrantc> for a quick and dirty example ...
| |
| 18:58 | there are lots of corner-cases that doesn't addres, but that's a starting point
| |
| 19:00 | jgee has joined IRC (jgee!~jgee@190.159.118.121) | |
| 19:00 | <spaced0ut> so if in my /etc/nbd-server/conf.d/config i have export /home/staticltspuser
| |
| 19:00 | i just would mount -t tmpfs tmpfs /home/staticltspuser
| |
| 19:01 | <alkisg> It's tmpfs OR nbd, not both
| |
| 19:01 | <spaced0ut> i'd expect mounting an nbd-export to look like mount ip:port:exportname
| |
| 19:02 | <alkisg> 2 different methods, unrelated between them
| |
| 19:02 | <spaced0ut> wait i see what vagrant is saying
| |
| 19:02 | gonna try that
| |
| 19:06 | <vagrantc> i was talking about something you could drop into /usr/share/ldm/rc.d/X00-mount-home or soemthing
| |
| 19:07 | or do it with FSTAB_MOUNT_HOME or INIT_COMMAND_MOUNT_HOME or soemthing like that
| |
| 19:07 | <spaced0ut> yep thats exactly what im doing
| |
| 19:07 | <vagrantc> doing it in ldm/rc.d/ would allow you to do things like reset it at each login
| |
| 19:08 | doing it in lts.conf wouldn't require updating the image
| |
| 19:08 | <spaced0ut> copied the servers /home/ltspuser directory to /tmp and am mount -t tmpfs /tmp/ltspuser /home
| |
| 19:09 | <vagrantc> you're getting too creative
| |
| 19:10 | i also may not be understanding the full context ... tmpfs would only be a good option if you can use an empty homedir
| |
| 19:10 | if you need it pre-populated with content, it might not be the way to go
| |
| 19:12 | <spaced0ut> gotcha gotcha
| |
| 19:16 | <vagrantc> i haven't experimented with alkisg's idea of using an NBD cow export, but of course it has many merits :)
| |
| 19:42 | jgee has left IRC (jgee!~jgee@190.159.118.121, Quit: The Lounge - https://thelounge.github.io) | |
| 19:50 | lucas_ has joined IRC (lucas_!~lucascast@200.141.207.18) | |
| 19:50 | lucascastro has left IRC (lucascastro!~lucascast@170.78.53.20, Remote host closed the connection) | |
| 19:50 | lucas_ has joined IRC (lucas_!~lucascast@200.141.207.18) | |
| 20:28 | lucas_ has left IRC (lucas_!~lucascast@200.141.207.18, Remote host closed the connection) | |
| 20:42 | Faith has left IRC (Faith!~Paty_@unaffiliated/faith, Quit: Leaving) | |
| 21:10 | jgee has joined IRC (jgee!~jgee@190.159.118.121) | |
| 21:41 | ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving) | |
| 22:23 | kjackal_v2 has joined IRC (kjackal_v2!~quassel@ppp-2-86-54-15.home.otenet.gr) | |
| 22:23 | kjackal has left IRC (kjackal!~quassel@2a02:587:3101:400:8c1b:8441:9019:ed7a, Ping timeout: 264 seconds) | |
| 22:48 | kjackal_v2 has left IRC (kjackal_v2!~quassel@ppp-2-86-54-15.home.otenet.gr, Ping timeout: 252 seconds) | |