[00:02] <RevertToType> this is gonna be effort as heck meh
[00:04] <sarnold> RevertToType: looks like chromium has a download locations thing, check about:settings, open advanced, and set the thing..
[00:04] <sarnold> RevertToType: then check where it is set in ~/.config/chromium* ...
[00:06] <RevertToType> yeah the problem is while i can set that since there's like an ext. to open office docs and whatnot we thought that was super useful given our typical use... so if someone has to save it's a right click save as situation for like 90% of the types of files we'd be using... and then it's gotta go to filechooser,
[00:06] <sarnold> you probably just need to pre-populate the save-as with something that they can actually write to
[00:06] <RevertToType>  i set the download location to /media/usb0 which is perfect for the auto-downloads but moreso when they're using say the office extension or a ... umm thingamabob that requires a right click i'd love it to just force the filechooser to either be bypassed altogether cause it's just auto-choosing /media/usb0 or the chooser shows just that one choice
[00:37] <patdk-lap> sarnold :)
[00:37] <patdk-lap> good in apparmor?
[00:38] <patdk-lap> I want a rule that just logs connections, just does like a warning, not block
[00:38] <patdk-lap> just so I know what user/program/...
[00:45] <sarnold> patdk-lap: hmm, that might be difficult. You might try to use "audit network," or "audit network tcp," in a  profile, but I'm not sure that'll work.
[00:46] <sarnold> patdk-lap: you could use auditd and log accept() syscalls; that would apply to all programs, without requiring apparmor profiles
[01:11] <patdk-lap> heh
[01:11] <patdk-lap> I would love to limit it to apparmor, as it would limit the insane amount of logging to the specific things I care about :)
[01:11] <patdk-lap> but if I have to
[01:16] <patdk-lap> audit network tcp connect, seems like it will do the trick
[01:21] <sarnold> patdk-lap: let me know if it works :)
[01:30] <patdk-lap> besides the connect part it does
[01:30] <patdk-lap> I think connect requires src/dst addresses
[01:30] <patdk-lap> manual is very fuzzy
[01:43] <patdk-lap> heh, connect doesn't work, says it needs v3.0
[01:43] <patdk-lap> and only v2.8
[01:43] <patdk-lap> and without it, I'm blowing up the kernel, logging stuff to fast
[01:44] <patdk-lap> audit_printk_skb: 43755 callbacks suppressed
[02:16] <sarnold> patdk-lap: ouch :)
[05:19] <lampshades> Hi there, I'm trying to provision an ubuntu server with chef and capistrano for a rails site, and Im having a really hard time finding good resources for this task
[08:16] <liecy> hello
[08:17] <lordievader> o/
[08:23] <silvervulcan> I have created my own apache ssl cert for my home server and am using the https address on port 443. I noticed that my server does DNS requests every time I access any page. When I use the regular port 80 address this doesn't happen. Is apache trying to look my home-made certificate in a public database?
[08:26] <lordievader> silvervulcan: I'd checkout what the DNS packets contain.
[08:26] <silvervulcan> I should have said self-signed certicate instead of home-made, that's what I meant.
[08:26] <silvervulcan> lordievader, how do I do that?
[08:26] <lordievader> silvervulcan: tcpdump/wireshark
[08:27] <silvervulcan> lordievader, will the output be in an easy to understand format or binary/hex?
[08:28] <lordievader> Wireshark will display it in an easy to understand format and in hex.
[09:59] <orogor> hi
[10:00] <orogor> anyone knows a good/simple tool for proviing a referential of the server inventory
[10:00] <jamespage> hallyn, zul: I jumped to a conclusion for https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1495895
[10:00] <jamespage> it may be wrong
[10:01] <jamespage> but it might not be
[10:01] <orogor> something that can allow to define the server as test/prod and a role like db/web ?
[10:14] <sysrex> orogor, you mean puppet / chef like
[10:21] <orogor> well no conf management; more like inventory
[10:21] <orogor> more like icmdb or glpi
[10:21] <orogor> but simpler
[10:22] <orogor> i am also looking for centralised ssh key management that supports the management server being offline/disconnected
[10:22] <orogor> and eventually a tool like keypass, but for multiuser
[10:23] <orogor> doesnt need to be the same app for everything
[10:32] <rbasak> teward: I just triaged bug 1494869 and noticed you were involved in the bug upstream. Perhaps you could help drive it? There's not enough for anything to be done in Ubuntu based on the report filed IMHO.
[11:54] <zul> jamespage: ive jumped to the same conclusion
[12:13] <sudhir__>  Hi guys.. I'm facing an issue while doing a net install of ubuntu server 14.4.2.  it says "configuring pkgsel fails with error code 127", can anybody help.??  I'm using preseed file for the installation.  I face this issue in select and install softwares step. in the screen i see ,running popcorn..tasksel..  cleaning up.. and then it just gives a red screen sayin installation step failed.. !!
[12:14] <sudhir__> guys any insights would be a great help.. :)
[12:25] <jamespage> zul, there is a 'configuration failed' message during configuration phase of the build
[12:26] <zul> jamespage: looking
[12:37] <zul> jamespage: can use qemu-img create to create a file with the rbd backend just to make sure
[12:41] <sudhir__> Hi guys.. I'm facing an issue while doing a net install of ubuntu server 14.4.2.  it says "configuring pkgsel fails with error code 127", can anybody help.??  I'm using preseed file for the installation.  I face this issue in select and install softwares step. in the screen i see ,running popcorn..tasksel..  cleaning up.. and then it just gives a red screen sayin installation step failed.. !!
[12:51] <zul> jamespage: its definenly built with rados/rbd support though
[12:56] <jpds> sudhir__: Sounds like your preseed file is broken?
[13:24] <jamespage> zul, why is it not linked against that then?
[13:24] <zul> im not sure
[13:25] <zul> but its building with -lrados
[13:28] <zul> jamespage, http://pastebin.ubuntu.com/12417537/
[13:49] <zul> jamespage: erm http://pastebin.ubuntu.com/12417627/
[13:50] <zul> jamespage: looks better now http://pastebin.ubuntu.com/12417636/
[13:51] <kgirthofer> if I want to pipe the output of a command into another command how do I do that
[13:51] <kgirthofer> i.e. if I want the output from "aws ec2 describe-instances --filter Name=tag:Name,Values=MYSERVER01 | grep "InstanceId" | awk '{print $2}' | tr -d '",'"
[13:52] <kgirthofer> to be piped into "aws ec2 reboot-instances --instance-ids $serveridhere
[13:52] <kgirthofer> how do it
[14:04] <KlausedSource> well you save it in a variable and use the variable in the command
[14:04] <KlausedSource> basically as you just did with $serveridhere
[15:27] <jamespage> zul, ok so which piece of the puzzle are we missing? is this a packaging fix, or a charm fix?
[15:28] <zul> jamespage: packging fix, rharper and hallyn are already working on it
[16:31] <teward> rbasak: which bug?
[16:31] <teward> oh
[16:31] <rbasak> teward: https://launchpad.net/bugs/1494869
[16:31] <teward> rbasak: What do you specifically need?
[16:31] <rbasak> teward: I don't need anything, but it looks like this needs to go upstream?
[16:32] <teward> rbasak: to where, OpenSSL?  or bind9?
[16:32] <rbasak> teward: it isn't clear to me.
[16:32] <teward> rbasak: nor me, the only reason I was involved was because their script failed
[16:32] <teward> beyond that, they're the SSL experts
[16:33] <rbasak> teward: I just thought you might care to drive it because of your upstream involvement. If you aren't so inclined, then that's no problem.
[16:33] <teward> rbasak: the only thing I can think of is an Ubuntuism / OpenSSLism where openssl_conf isn't defined as an env var
[16:33] <teward> rbasak: yeah i'm not particularly driven about that
[16:33] <teward> since the issue appears to have autoresolved itself over time :/
[16:33] <teward> my guess is because a fix was made in the script
[16:33] <rbasak> teward: could be. I don't feel that the bug is reported well enough to know.
[16:36] <teward> rbasak: TBH I don't think it warrants fixing, unless someone cares enough to.  I know that 15.10+ is not affected, and I"m more concerned currently with the state of 15.10+ than I am a weird bug that had a workaround applied as part of the script that Dirk was developing
[16:36] <teward> i'm also more concerned about why openssl ftbfs on my rpi, but that's a different issue :)
[16:36] <rbasak> Yeah that's fine. I commented on the bug just because otherwise reporters have the expectation that we're supposed to be doing something but aren't, rather than the bug not being actionable.
[16:36] <teward> and i'm about 3 weeks behind on the PPA packages lol
[16:54] <teward> sarnold: ping, if you're around.  If not i'll bug you later
[17:16] <lochlann> I have done so many searches on google about the proper way to automount ntfs drives and have come up with so many "different" answers.  Please someone help me with the "proper" way to do this.  I want to automount ALL ntfs drives so they are RW by all users on the local machine and accross the network.
[17:18] <teward> lochlann: are the NTFS drives on a file share server?
[17:18] <teward> or are they actual drives/partitions
[17:18] <lochlann> some are internal Sata and some are external USB ... It's for a media server
[17:20] <teward> lochlann: /etc/fstab entries for each partition/device to mount it rw, anonymous FTP internal to the rest of your network.  Or use an actual media streaming solution rather than sharing partitions and files and such
[17:21] <lochlann> there is 5 internal hdd and 6 external drives. I have a media streaming solution in use *Plex* and this is the Plex Server where all the drives are to be mounted.
[17:22] <teward> lochlann: /etc/fstab entries then
[17:22] <teward> and either Plex to share from, or ftp/sftp internal only for direct files access
[17:22] <teward> (internal network that is, i.e. your LAN)
[17:23] <lochlann> teaward: yes I know fstab... my issue as stated is I am finding *via google* way to many different ways of adding it to fstab. I have tried some and they are not working properly. I'm looking for the proper way to "add" it to fstab
[17:23] <teward> you didn't say that lol
[17:23] <teward> read the manpage perhaps, it explains NTFS mount options
[17:26] <teward> This is how I mount an NTFS USB Hard disk: UUID=fwafwefwafweffwaew  /media/DIRTY_150G    ntfs-3g    defaults,locale=en_US.utf8,windows_names,umask=0000,nobootwait,uid=1000,gid=1000,user   0       0
[17:27] <teward> uid=1000, gid=1000 is my user id, umask=0000 ends up giving chmod 777-like perms (read/write)
[17:27] <teward> (to all)
[17:27] <lochlann> what is this windows_names thing in yours? I am guessing it has something to do with windows and not needed by me? I have no windows machines in this house
[17:28] <teward> lochlann: as i said it's an NTFS drive and I want cross-compatibility at times
[17:28] <teward> lochlann: that's just one example
[17:28] <teward> as i said, you need to read the manpages for mount options for ntfs
[17:29] <teward> lochlann: we don't know your requirements.  which is why you need to read the manpages.
[17:29] <lochlann> what manpage are you speaking of?    mount -h says nothing about ntfs
[17:30] <teward> that's not a manpage
[17:30] <teward> type `man mount`
[17:31] <teward> and then read through the general fs-nonspecific section, as well as the ntfs options
[17:48] <lochlann> teward: so then, from  what I just read, this should work -->   UUID=1293078mnbndf89      /media/VideoDrive01      ntfs-3g     auto,defaults     0     0
[17:50] <tarpman> careful. ntfs and ntfs-3g are different things and have different options. mount(8) has ntfs options, ntfs-3g options are in ntfs-3g(8)
[17:50] <tarpman> ntfs-3g has many more options than ntfs
[17:52] <teward> ^ that
[17:53] <lochlann> I really just want it to 1. automount in the off-chance of a system reboot. 2. be accessible via Plex for all. 3. be rw so new media can be added from other machines on the network.
[17:53] <teward> lochlann: ideally if you don't know what you're doing, though, you shouldn't be messing with mounting things.
[17:53] <teward> (in my opinion)
[17:55] <lochlann> teward: it's not about messing with it..... I just need to do it once...
[17:55] <teward> lochlann: `man ntfs-3g` may be of interest moreso
[17:55] <teward> and yes it's 'messing' until you get it right
[17:55] <teward> you also haven't said how data's being written to it from remote sources
[17:55] <teward> FTP?  SFTP?  Windows file shares (Samba)?  etc.
[17:56] <teward> that also has a factor too
[17:56] <tarpman> lochlann: just curious: if you have no windows machines in your house, why are you bothering with ntfs?
[17:57] <teward> tarpman: external USB media, perhaps preformatted
[17:57] <teward> [2015-09-15 13:21:45] <lochlann> there is 5 internal hdd and 6 external drives. I have a media streaming solution in use *Plex* and this is the Plex Server where all the drives are to be mounted.
[17:57] <teward> pretty certain preformatted
[17:57] <lochlann> tarpman: from what I read, ntfs was the way to go for the size of the drives. 11 drives all 4TB each. plus the usb drives came this way when bought
[17:58] <teward> if your backend is *nix, then ext4 works
[17:59] <lochlann> plus there is, albeit not very often, 1 windows laptop that connects when family comes to visit.
[17:59] <tarpman> the stuff you read probably assumed you need compatibility with windows or mac
[18:00] <tarpman> I format all my usb drives as ext4 because I find dealing with the permissions-translation stuff (unix to ntfs and back) to be a massive pain
[18:00] <lochlann> although I don't see the need for ntfs if all they are using is Plex.... but everything I read says this
[18:00] <tarpman> anyway, not advice per se, was just curious about the rationale
[18:00] <teward> lochlann: if they aren't connecting the drives themselves then NTFS is not a requirement
[18:00] <tarpman> yeah, only the machine the usb drives are actually plugged into cares about the filesystem. over the network, it's irrelevant
[18:00] <teward> if they're always connected to your media server, then filesystem matters only to the server
[18:00] <teward> and ext4 is more... linux friendly, and flexible, I think
[18:00] <teward> over the network it's irrelevant
[18:01] <lochlann> and now I think it is too late....  all 11 drives have media on them now
[18:01] <RoyK> xfs is probably better unless you want to shrink the fs
[18:02] <RoyK> lochlann: how large drives? what raid level?
[18:02] <lochlann> RoyK: no raid... all drives are 4TB each
[18:02] <RoyK> lochlann: that's lunacy
[18:02] <teward> RoyK: they're also not completely internal according to the user
[18:02] <RoyK> lochlann: with 11 drives, create a raid-6 on top of them
[18:03] <lochlann> ^^
[18:03] <teward> RoyK: they want all 11 to automount and have the media on them available r/w to the system and to external 'file dropping' into the server probably
[18:03] <RoyK> oh - lots of usb drives around?
[18:03] <teward> RoyK: [2015-09-15 13:21:45] <lochlann> there is 5 internal hdd and 6 external drives. I have a media streaming solution in use *Plex* and this is the Plex Server where all the drives are to be mounted.
[18:03] <lochlann> 5 internal 6 external
[18:03] <RoyK> get it
[18:03]  * RoyK would prefer a large raidset
[18:03] <lochlann> technically there is 6 internal, but one houses Ubuntu
[18:04] <teward> RoyK: wouldn't the RAIDing nuke the data on the disks already?
[18:04] <sarnold> lochlann: yikes, this sounds terrifying..
[18:04] <sarnold> that's a lot of singlepoints of failure..
[18:04] <teward> sarnold: there you are
[18:04] <sarnold> heya teward :)
[18:04] <lochlann> and here I thought this would be simple.... lol....  as it is it took my wife and I almost a year to take the media off over 1,200 DVD's
[18:04] <tarpman> teward: in theory that's not a problem because one has up-to-date backups...
[18:04] <teward> tarpman: in theory
[18:05] <tarpman> ^_^
[18:05] <teward> tarpman: in practice, people making in-home media servers aren't up to theory
[18:05] <RoyK> teward: just empty one drive, create a broken r5 on that drive, create lvm on top, mkfs on that, copy data, empty another, extend etc
[18:06] <sarnold> pff too easy to screw that up; buy a nice new system with a dozen drives, set up a raidz3 on them, start copying, and pray to all the dieties of your choice that those usb drives last long enough to copy everything off
[18:06] <lochlann> sarnold: all the drives are 1year old or less :)
[18:08] <bekks> lochlann: drives can be dead on arrival, their age basically doesnt mean much.
[18:08] <sarnold> lochlann: even worse, some haven't even been alive long enough to be stressed :)
[18:09] <RoyK> sarnold: well, you can't extend zfs vdevs
[18:12] <lochlann> trial and error I suppose. I will start with -->   UUID=1293078mnbndf89      /media/VideoDrive01      ntfs-3g     auto,defaults     0     0 and see if it works for me
[18:12] <sarnold> RoyK: true :(
[18:13] <tarpman> lochlann: i'd suggest 0 2 instead of 0 0
[18:15] <tarpman> oh, nevermind. there's no fsck.ntfs-3g anyway...
[18:16] <lochlann> wouldn't it be bad to have it utilize parallelism on 11 drives though?
[18:18] <tarpman> I can't think why
[18:35] <lochlann> took a while to add all those, but so far all seems good.
[18:36] <RoyK> just don't use ntfs on linux
[18:36] <lochlann> yea, all future drives added with be EXT4
[18:37] <lochlann> we still have approx 900 DVD's to convert
[18:37] <lochlann> my wife is a bloody dvd fanatic
[18:38] <lochlann> thank the gods and goddesses for .mkv .....
[18:38] <lochlann> lol
[18:39] <jrwren> lochlann: not 11 drive, but I have 5 drives in an LVM VG of ~11TB with larget LV being 5TB, but each grows as I add disks.
[18:40] <jrwren> lochlann: I'd NEVER use NTFS for any of it. It is ext4. I use some smaller LV to experiment with btrfs, but I'm overly cautious and do not trust it just yet.
[18:40] <lochlann> I suspect we will be adding at least a few more drives....  only got about 7TB left of space from all 11 4TB drives... and still over 900 dvds to convert ......
[18:41] <lochlann> maybe I will take the new drives as an opportunity to convert the existing ones to ext4
[18:44] <jrwren> lochlann: it is probably worth adding them all to some kind of pool with parity drives so that you can suffer a disk outage. raid6 or btrfs or something.
[18:59] <lochlann> jrwren: no point in me doing that honestly....  That's another 11 drives I would have to buy plus anymore that I add in the future.....  This is only a media server... Should 1 drive fail I will just reconvert the dvd's that were on that drive. much cheaper option
[19:03] <jrwren> lochlann: no, you strongly misunderstand. I wrote parity on purpose.
[19:03] <jrwren> lochlann: it is only cheaper if your time is free. My time is very expensive and very valuable to me.
[19:05] <freezevee> I am trying to do sudo passenger-install-nginx-module while I have installed ruby 2.2 via rvm and checked with ruby -v but it keeps installing for the 1.9.3 version. What am I doing wrong ?
[19:05] <jrwren> freezevee: sudo resets your path. try sudo -E
[19:08] <lochlann> jrwren: that involves buying more drives does it not? If so that means buying another 11 4TB drives just for what I have now. Further, my machine only has 4 more USB ports available which would not accommodate another 11 drives
[19:09] <jrwren> lochlann: no, parity means some, not all. Yes you would sacrifice some space. Its ok. I was only suggesting. We may value time differently.
[19:10] <freezevee> jrwren: should I install without the sudo ? nice idea
[19:10] <lochlann> jrwren: not a bad suggestion, but financially for me it is more worth the time to reconvert
[19:41] <dft> any incrontab afficionados here?
[19:42] <dft> having issues trying to pinpoint execution of a python script from incrontab
[19:42] <dft> I can see that the inotify signal is being caught and incrond is executing my python script, but the desired result of mailing me the results is not happening.
[19:43] <dft> when I run the python script from the users shell, all is well.
[19:43] <sarnold> dft: how are you checking the email delivery?
[19:45] <dft> sarnold: email delivery works fine when I run the script from my shell
[19:45] <dft> as in it reaches my inbox
[19:46] <dft> sarnold: before tracing that, I don't see the script executed in the users shell under top at all when I would normally expect it.
[19:46] <dft> I tried sourcing .bashrc for the user account prior to executing the python but I get other errors using that method
[19:47] <sarnold> dft: so, the script itself is doing the mailing? does it make assumptions about the PATH or other environment variables that might be set in your shell but not set in incron?
[19:47] <sarnold> dft: top is probabilistic, it's easy for it to miss executions; fatrace or "execsnoop" can help there...
[19:48] <sarnold> dft: see http://www.brendangregg.com/blog/2014-07-28/execsnoop-for-linux.html for more information on execsnoop; awesome blog overall, it's well worth a few days of reading :)
[19:48] <dft> I'll check those out.
[19:48] <dft> ty
[19:48] <dft> but as far as assumptions, I think I've covered them all.
[19:49] <dft> shebang points to /usr/bin/python, then just importing all my required classes
[19:50] <dft> although I'll have to look into os.path.dirname(os.path.abspath(__file__)
[19:50] <dft> I'm not sure if that pulls from env vars
[19:51] <dft> otherwise, any other filesystem path is explicitly defined
[19:51] <sarnold> can you pastebin the incrontab file?
[19:52] <RevertToType> i don't have network-manager installed would this be a reason why my wifi keeps dropping or is that just a suite of tools to make things easier?
[19:52] <dft> sarnold: standby
[19:52] <dft> it's the last entry that's problematic atm
[19:53] <dft> sarnold: https://gist.github.com/anonymous/d1a540bab1f8f5bb338d
[19:54] <sarnold> dft: that . is a little funny but it shouldn't cause problems to the kernel...
[19:55] <sarnold> dft: is there a newline at the end of the file? some programs require a newline at the end, and some editors don't put one there
[19:55] <dft> it's never been an issue with the first 4 entries
[19:55] <dft> sarnold: let's find out...
[19:55] <sarnold> dft: are the modes on /home/elimg/scripts/./funcCollectDefectImages.py  set correctly?
[19:55] <dft> yes
[19:55] <dft> 755
[19:57] <dft> I think I may have found the issue
[19:57] <dft> your comment about assumptions may have sparked a notion
[19:57] <dft> I have some logfile writing functions but the log file path is not fully qualified
[19:57] <dft> brb
[20:03] <dft> BAM
[20:03] <dft> sarnold: ty
[20:03] <dft> needed someone to bounce this off of
[20:03] <sarnold> dft: woot :)
[20:04] <dft> :facepalm: I've been chasing this all day
[20:04] <dft> so future reference
[20:05] <dft> don't use os.path.dirname(os.path.abspath(__file__) to getcwd
[20:05] <dft> at least when calling your py from cron or incron
[20:06] <dft> declare a wrkdir='<home>' and use os.chdir(wrkdir)
[20:15] <sarnold> dft: funny, I would have expected that to work
[20:16] <dft> sarnold: as you first stated....assumptions....
[20:16] <dft> I thought so too
[20:17] <dft> but as it turns out I've bounced between os.path.dirname and my 2nd in some of these scripts.
[20:18] <dft> after some googling I found some references to issues with incrontab and CentOS not providing a full shell when executing the referenced script
[20:18] <dft> so that's when I tried sourcing my .bashrc...however that just made matters worse.
[20:20] <sarnold> yeah, things that make a shell friendly for humans often make it less friendly for programs ;) hehe
[20:21] <dft> oh well, in the end I get to level up ..haha
[20:22]  * dft +1
[20:26] <sarnold> :)
[22:22] <RevertToType> so i've narrowed down at least kind of a seeming culprit to my wireless drops (drops around 1800/30 minutes) and it's something to do with dhclient apparently... once it drops the retry isn't kicking in either so even after five minutes...  so it's like dhclient isn't renewing or even running apparently
[22:25] <RevertToType> how do i fix this awfulness
[22:26] <genii> RevertToType: Is the wireless adapter running on the USB bus? It may be going into power saving mode
[22:26] <sarnold> genii: that's genius :)
[22:27] <RevertToType> genii it's likely ... it's a netbook like half the internals are technically on the usbbus
[22:27] <RevertToType> i thought i killed out all powersaving...
[22:28] <RevertToType> also i don't have acpi-tools or network-manager installed at all
[22:29] <RevertToType> where would i futz that one maybe i did it wrong (power saving stuff)
[22:30] <genii> Apologies on lag, just getting ready to leave work here...
[22:30] <genii> sarnold: I had a broadcom like that, drove me nuts
[22:30] <RevertToType> np grateful for any things
[22:31] <genii> RevertToType: Does it show up on hte: lsusb   ..command?
[22:31] <RevertToType> umm the only named thing i recognize aside from linux foundation..... is the camera
[22:33] <RevertToType> so maybe it's not on usb
[22:33] <genii> RevertToType: Could you please pastebin the output of: lsusb   and: sudo lspci    and: sudo lshw -short     please
[22:34] <RevertToType> oh it's on pci
[22:34] <sarnold> genii: it reminds me of a usb hard drive I used to use for backups. It'd go to sleep and the mount failed and all kinds of hell would break loose and then it'd require fscking every time I wanted to use it. Horrible annoyance.
[22:35] <genii> So in this case it looks like some other issue, unfortunately
[22:35] <RevertToType> i mean i think the culprit has something to do with dhclient not doing what it's supposed to
[22:36] <RevertToType> cause when it drops
[22:36] <genii> RevertToType: Did you check on the place it's getting it's lease from to see what the default lease length is?
[22:36] <RevertToType> it's still saying it's connected to the network/associated and the ip address drops out of ifconfig after a few minutes
[22:37] <RevertToType> i have no access to the network settings or even the router (this is for a small library and the IT is just the county folks... if something goes wrong and it's not windows they tell us to leave)
[22:38] <sarnold> RevertToType: check dmesg, auditd logs, etc.. there might be some clues hidden somewhere you may not expect
[22:38] <RevertToType> and while somewhat windows/mac savvy i'm like just one step above totally clueless on linux (so enough knowledge to do damage)
[22:39] <RevertToType> all i ever see is a deauthenticate reason 108; unknown everything else looks like it's getting orders from the accesspoint/router/whatever just fine
[22:39] <sarnold> wpa2 vs wep or something?
[22:40] <RevertToType> wpa2-personal so form wht i can tell wpa-psk ccmp
[22:40] <RevertToType> hidden ssid
[22:40] <RevertToType> when i finally got wireless working i had to make a systemd service to run "dhclient" at startup
[22:40] <RevertToType> cause it wasn't pulling down an ip
[22:41] <genii> hidden SSID can be problemmatic
[22:44] <genii> Are you using a wpa_supplicant.conf file ?
[22:45]  * genii ignores the phone ringing in his office and makes more coffee
[22:48] <RevertToType> genii yup
[22:48] <RevertToType> it's set to scan=1 so it picks up and connects just fine
[22:48] <RevertToType> after 30 minutes it drops though
[22:48] <RevertToType> yet running dhclient -r / dhclient or just dhclient picks it back up
[22:49] <RevertToType> so i assumed the retry was just too long and the lease was just generally expiring so i waited the 300 seconds and no dice still down
[22:50] <RevertToType> i had to run dhclient manually at boot every time to get an ip and be able to do normal internet things
[22:51] <genii> RevertToType: Might want to see if the dhcp server accepts request for specific lease time by twiddling with the: send dhcp-lease-time 3600;    ..line in the /etc/dhcp/dhclient.conf file
[22:52] <RevertToType> dhclient is a one shot... i don't have dhcpcd installed... that might be the issue?
[22:52] <RevertToType> or run dhclient -w?
[22:52] <RevertToType> cause i need something ongoing when the lease expires?
[22:53] <RevertToType> am i just that cluelessly stupid?
[22:53] <RevertToType> i am i think
[22:55] <genii> You don't need dhcpd, that one is for assigning IPs to clients :)
[22:55] <RevertToType> dhcpcd
[22:55] <RevertToType> not dhcpd
[22:57] <genii> Ah, misread. Possibly, it does try to renew automatically. But the regular dhclient should too.
[22:58] <RevertToType> hrm well that was a failure of epic proportions...
[22:58] <sarnold> I wonder if systemd is killing the dhclient unexpectedly?
[22:58] <RevertToType> i just installed and tried it gave a whole minute of connection and then was like nope
[22:59] <sarnold> you might want to do some ghetto-troubleshooting like having a while true ; do sleep 1; ps auxw >> jobs ; done   -- and when the IP drops, see if dhclient was running at the time..
[22:59] <genii> ewww
[22:59] <RevertToType> yeah i was really thinking some hacky thing like that ugh
[22:59] <RevertToType> really didn't want to
[23:02] <genii> RevertToType: I have to leave, but I'm around weekdays 9:30am-5pm EDT  ( later tonight than usual) , if it's not solved tonight we could poke at it again tomorrow
[23:02] <RevertToType> ok... i'm really grateful for all of your help folks
[23:02] <RevertToType> like honestly
[23:04] <RevertToType> removed dhclient from the systemd services and am gonna test with dhcpcd to see how that goes just in case
[23:05] <RevertToType> cause it was for sure not playing well together maybe with a bit-o-luck that was all i needed to do... i'm sure i'll be back in 30 minutes tho :V
[23:07] <teward> thought dhcpd was a dhcp server/daemon, not a dhcp client
[23:08] <genii> !info dhcpcd
[23:08] <teward> ah
[23:08]  * teward misread :)
[23:08] <sarnold> not at all confusing is it? :)
[23:09]  * genii wanders off
[23:11] <teward> sarnold: heh