| drab | mmmh is there a way to reset a container to its original image? kinda like restoring froma snap, except that its the original image | 02:44 |
|---|---|---|
| drab | I don't wanna destroy the container because it has a whole bunch of configs I wanna preserve, altho I could sae those and reapply them if there's no other way | 02:45 |
| drab | with LXD I mean | 02:45 |
| sarnold | how would you pick which bits to save and which to roll back? | 02:46 |
| drab | I was thinking more along the lines of automating what I'd do manually, save config, destroy, relaunch, reapply config | 02:47 |
| sarnold | ah | 02:47 |
| sarnold | sort of like juju or ansible or salt or puppet or chef or whatever's hot -this- week? :) | 02:48 |
| drab | not like a real snap/restore operation, but make it transparent as if it was | 02:48 |
| drab | yeah, that's acually what I was about to do... with ansible | 02:48 |
| drab | all my other hosts are on that, so just getting the hang of it to manage containers too | 02:48 |
| drab | but I've yet to crack that nut so I was looking for shortcuts while experimenting | 02:48 |
| drab | but it's a good time as any I guess to figure it out | 02:49 |
| st3fann | Hi there, I am making an OpenVZ template from a debian minimal install, and I noticed that all of the packages I installed are stored in .deb files at /var/cache/apt, and I wanted to know if these are safe to delete | 03:46 |
| st3fann | I did run apt-get clean and apt-get autoclean, but this did not seem to remove the files. | 03:46 |
| sarnold | st3fann: yes, those are safe to delete | 03:53 |
| sarnold | I'm surprised apt-get clean didn't do the trick | 03:53 |
| sarnold | but feel free to delete them all | 03:54 |
| st3fann | me too | 03:54 |
| st3fann | went to separate threads, and some said if one didnt clear it the other should | 03:54 |
| st3fann | do I just remove the entire directory sarnold? | 03:54 |
| sarnold | st3fann: I'd recommend just deleting the files | 03:54 |
| sarnold | probably the directory would be recreated with proper permissions if needed but I never like finding out the hard way when things don't work as expected :( | 03:54 |
| st3fann | Sounds good - any way to restart apt after to see if it's okay? I might just move a file to see if it causes any issues | 03:55 |
| st3fann | yep was just thinking the same thing :) | 03:55 |
| sarnold | apt isn't a long-running service; just apt-get update && apt-get -u dist-upgrade or whatever your favorite finger-macros are :) | 03:55 |
| drab | ok, that wasn't so bad, automation done, pretty sweet actually | 03:56 |
| st3fann | sarnold: awesome, thank you so much! Ill try it out right now | 03:56 |
| drab | and half of the hr was waiting for images to redownload since I managed to wipe my local test repo... | 03:56 |
| sarnold | drab: d'oh :/ | 03:57 |
| sarnold | st3fann: excellent :) have fun | 03:57 |
| sarnold | drab: https://xkcd.com/303/ (not the same, but close) | 03:57 |
| drab | lol, I thought of that actually, and it actually did provide me with an excuse for a break to practice writing characters | 03:58 |
| sarnold | not bad :) | 04:00 |
| drab | mmmh | 04:44 |
| drab | a minute and a half to stop a blank container, that doesn't seem right | 04:44 |
| drab | is there a way to figure out what's going on in cases like this? | 04:44 |
| st3fann | does anyone here have any experience with solusvm/openvz? | 04:49 |
| lynorian | !anyone | 04:49 |
| st3fann | lynorian: ? | 04:51 |
| lynorian | can you ask your question instead of saying anyone have expirence with solusvm/openvz | 04:51 |
| lynorian | I unfortanetly do not | 04:51 |
| st3fann | I have created a template from a vm that i had running on solusvm, and when I created a new VM, it is not able to boot | 04:52 |
| st3fann | I don't see any errors, it tells me to look in: dmesg, and I have this error: EXT4-fs (ploop14464p1): loaded balloon from 12 (4096 blocks) | 04:52 |
| st3fann | thanks | 04:52 |
| === aw17 is now known as AW17 | ||
| cpaelzer | good morning | 06:47 |
| lordievader | Good morning | 08:22 |
| danpawlik | coreycb: Hello. I don't know if it is good, but nova-common from cloud-archive ocata "branch" is failing on installation | 09:56 |
| sonu_nk | hi i installed phantomjs on my ubuntu server.. not i want to check where it is intalled .. how can do ? | 12:04 |
| cpaelzer | sonu_nk: with "dpkg -l <pkg>" you can list files owned by a package | 12:05 |
| cpaelzer | sonu_nk: sorry -L | 12:05 |
| cpaelzer | sonu_nk: lower case lists install state andversion, upper case is content (--listfiles) | 12:05 |
| sonu_nk | i installed using this command sudo apt-get install phantomjs | 12:05 |
| sonu_nk | thankyou working cpaelzer | 12:08 |
| sonu_nk | also one more thing .. using this command sudo apt-get install phantomjs ubuntu installed its lower version .. Why ??? but its have its new version on phantomjs site | 12:08 |
| cpaelzer | sonu_nk: this is the version in the Ubuntu Archive for your Ubuntu release | 12:09 |
| sonu_nk | ubuntu server installed 1.9.0 version but there is 2.1.1... | 12:10 |
| sonu_nk | cpaelzer, so means 1.9.0 is good and tested | 12:10 |
| sonu_nk | right | 12:10 |
| cpaelzer | you are likely on Trusty | 12:10 |
| cpaelzer | which is good, but to admit a 3 year old release | 12:10 |
| cpaelzer | there 1.9 was current | 12:10 |
| cpaelzer | on Later Ubuntu Releases (>=Xenial) you'd get 2.1.1 | 12:10 |
| === costingalan is now known as cgalan | ||
| coreycb | danpawlik, in what way does it fail to install? | 12:36 |
| coreycb | danpawlik, fwiw i just installed nova-common from the ocata cloud archive successfully on a new xenial machine | 12:39 |
| danpawlik | coreycb: strange | 12:40 |
| danpawlik | coreycb: I will paste you error log | 12:40 |
| danpawlik | coreycb: http://paste.openstack.org/show/8slHZRzdvr48sW3wThMt/ | 12:41 |
| danpawlik | coreycb: wait, This error is normall because I have sqllite parameter in nova.conf | 12:41 |
| danpawlik | but there was also onother error that it want to execute command which was changed | 12:42 |
| danpawlik | in ocata | 12:42 |
| danpawlik | let me check one more time on clean host | 12:42 |
| coreycb | danpawlik, ah right, the package will try to sync the db if you are using sqlite in nova.conf | 12:42 |
| danpawlik | IMHO there should be condition for that... or? | 12:43 |
| coreycb | danpawlik, yeah i don't think that accounds for the api database | 12:43 |
| coreycb | danpawlik, so that needs to be fixed | 12:43 |
| === tinwood_swap is now known as tinwood | ||
| danpawlik | it just irritates me when I would like to deploy host "all in one" using puppet ;P | 12:44 |
| coreycb | danpawlik, we might consider removing that sqlite bit since it's not meant for production use | 12:45 |
| danpawlik | coreycb: if someone will test Openstack he/she must use something "better" :p | 12:47 |
| danpawlik | It strange that Openstack developers doesn't drop support for sqllite | 12:47 |
| danpawlik | IMHO | 12:47 |
| coreycb | danpawlik, i opened a bug for that: https://bugs.launchpad.net/ubuntu/+source/nova/+bug/1671078 | 12:48 |
| ubottu | Launchpad bug 1671078 in nova (Ubuntu) "nova sqlite db migration failing for ocata" [Undecided,New] | 12:48 |
| danpawlik | coreycb: nice, thanks | 12:49 |
| frickler | oh, who had the funny idea of changing the group name libvirtd to libvirt for Ocata? how do you handle that for upgrades? | 13:05 |
| cpaelzer | frickler: I think that was done post xenial to align with Debian | 13:12 |
| cpaelzer | frickler: upgrades usually deal with this by dupping the gid | 13:12 |
| cpaelzer | frickler: if there is nothing it adds libvirt (as it formerly added libvirtd) | 13:13 |
| cpaelzer | frickler: if there is already a libvirtd it creates a new group libvirt with the same gid | 13:13 |
| cpaelzer | frickler: that way old and new names are synonyms | 13:13 |
| cpaelzer | I'm even more confused by libvirt vs libvirt-qemu groups sometimes | 13:15 |
| cpaelzer | frickler: let me know if that explains (a bit) | 13:15 |
| frickler | cpaelzer: o.k., thanks, now I just need to teach that to our cookbooks ;) | 13:18 |
| coreycb | jamespage, zul: do you know if ocata-proposed has been tested? it'd be nice to get qemu promoted for bug 1667033. | 14:02 |
| ubottu | bug 1667033 in qemu (Ubuntu) "nova instance console log empty" [High,Fix released] https://launchpad.net/bugs/1667033 | 14:02 |
| jamespage | coreycb: i've not done so - did the promotions late yesterday | 14:06 |
| coreycb | jamespage, ok. let me kick off some tests. it looks like kolla is tripping over it in the gate. | 14:06 |
| coreycb | danpawlik, It sounds like what you're hitting is similar to this so you are likely missing a step prior to installing ocata nova-common: https://bugs.launchpad.net/tripleo/+bug/1656791 | 14:29 |
| ubottu | Launchpad bug 1656791 in tripleo "Mitaka to Newton undercloud upgrade needs "nova db online_data_migrations" after the undercloud upgrade" [High,In progress] | 14:29 |
| Genk1 | Hello | 14:34 |
| Genk1 | Every mail client I use "Evolution, Thunderbird, mutt..." is based on DNS to resolve domain names | 14:34 |
| Genk1 | is there a way to tell the clients to use /etc/hosts instead | 14:35 |
| Genk1 | ? | 14:35 |
| Genk1 | I have a local server for test named : mail.example.com | 14:35 |
| Genk1 | but the client doesn't ask the hosts file instead he use the DNS frist | 14:36 |
| danpawlik | coreycb: another beautiful log file: http://paste.openstack.org/show/47mfXRuNNEbt9OZ0q1IJ/ | 14:46 |
| odc | Genk1: isn't that what /etc/nsswitch.conf is for? | 14:49 |
| coreycb | danpawlik, we'll need some context on that, ceph-mds installs ok for me on xenial | 14:50 |
| danpawlik | dpkg -l | grep ceph | 14:50 |
| danpawlik | ii ceph 10.2.3-1xenial amd64 distributed storage and file system | 14:50 |
| danpawlik | ii ceph-base 10.2.3-1xenial amd64 common ceph daemon libraries and management tools | 14:50 |
| danpawlik | ii ceph-common 10.2.3-1xenial amd64 common utilities to mount and interact with a ceph storage cluster | 14:50 |
| danpawlik | ii ceph-fs-common 10.2.3-1xenial amd64 common utilities to mount and interact with a ceph file system | 14:50 |
| danpawlik | ii ceph-fuse 10.2.3-1xenial amd64 FUSE-based client for the Ceph distributed file system | 14:50 |
| danpawlik | ii ceph-mds 10.2.3-1xenial amd64 metadata server for the ceph distributed file system | 14:50 |
| danpawlik | ii ceph-mon 10.2.3-1xenial amd64 monitor server for the ceph storage system | 14:50 |
| danpawlik | ii ceph-osd 10.2.3-1xenial amd64 OSD server for the ceph storage system | 14:50 |
| Genk1 | odc, I already have : hosts: files mdns4_minimal [NOTFOUND=return] dns | 14:50 |
| danpawlik | ii libcephfs1 10.2.3-1xenial amd64 Ceph distributed file system client library | 14:50 |
| danpawlik | ii python-ceph 10.2.3-1xenial amd64 Meta-package for python libraries for the Ceph libraries | 14:50 |
| danpawlik | ii python-cephfs 10.2.3-1xenial amd64 Python libraries for the Ceph libcephfs library | 14:50 |
| Genk1 | so it means to look for hosts file first | 14:51 |
| danpawlik | coreycb: Reading package lists... Done | 14:51 |
| danpawlik | Building dependency tree | 14:51 |
| danpawlik | Reading state information... Done | 14:51 |
| danpawlik | The following packages were automatically installed and are no longer required: | 14:51 |
| danpawlik | liblttng-ust-ctl2 liblttng-ust0 liburcu4 | 14:51 |
| danpawlik | Use 'apt autoremove' to remove them. | 14:51 |
| danpawlik | The following additional packages will be installed: | 14:51 |
| danpawlik | ceph-common ceph-mds libcephfs1 librados2 libradosstriper1 librbd1 librgw2 python-ceph python-cephfs python-rados python-rbd | 14:51 |
| danpawlik | The following packages will be REMOVED: | 14:51 |
| danpawlik | ceph-base ceph-mon ceph-osd | 14:51 |
| danpawlik | The following packages will be upgraded: | 14:51 |
| danpawlik | ceph ceph-common ceph-mds libcephfs1 librados2 libradosstriper1 librbd1 librgw2 python-ceph python-cephfs python-rados python-rbd | 14:51 |
| danpawlik | 12 upgraded, 0 newly installed, 3 to remove and 58 not upgraded. | 14:51 |
| danpawlik | 2 not fully installed or removed. | 14:51 |
| danpawlik | Need to get 0 B/40.7 MB of archives. | 14:51 |
| danpawlik | After this operation, 327 MB disk space will be freed. | 14:51 |
| danpawlik | Do you want to continue? [Y/n] | 14:51 |
| Genk1 | danpawlik, ? | 14:51 |
| danpawlik | (Reading database ... 77997 files and directories currently installed.) | 14:51 |
| danpawlik | Preparing to unpack .../ceph-mds_10.2.5-0ubuntu0.16.04.1_amd64.deb ... | 14:51 |
| danpawlik | Unpacking ceph-mds (10.2.5-0ubuntu0.16.04.1) over (10.2.3-1xenial) ... | 14:51 |
| danpawlik | dpkg: error processing archive /var/cache/apt/archives/ceph-mds_10.2.5-0ubuntu0.16.04.1_amd64.deb (--unpack): | 14:51 |
| danpawlik | trying to overwrite '/usr/bin/cephfs-table-tool', which is also in package ceph-common 10.2.3-1xenial | 14:51 |
| danpawlik | dpkg-deb: error: subprocess paste was killed by signal (Broken pipe) | 14:51 |
| danpawlik | Processing triggers for ureadahead (0.100.0-19) ... | 14:51 |
| danpawlik | Processing triggers for man-db (2.7.5-1) ... | 14:52 |
| danpawlik | Errors were encountered while processing: | 14:52 |
| danpawlik | /var/cache/apt/archives/ceph-mds_10.2.5-0ubuntu0.16.04.1_amd64.deb | 14:52 |
| danpawlik | E: Sub-process /usr/bin/dpkg returned an error code (1) | 14:52 |
| danpawlik | I will remove all cepch packages, install one more time 10.2.3 and upgrade to 10.2.5 | 14:52 |
| danpawlik | Genk1: sorry for spaming | 14:52 |
| danpawlik | I should paste it on pastebin | 14:52 |
| coreycb | danpawlik, next time please share a link to a pastebin | 14:52 |
| danpawlik | sorry | 14:52 |
| odc | wow | 14:52 |
| odc | Genk1: i don't know then :( | 14:52 |
| danpawlik | coreycb: http://paste.openstack.org/show/IF4qthaWuncT8tiSMkR6/ | 14:52 |
| coreycb | danpawlik, what is the source of those packages? they don't appear to be coming from the ubuntu archive. | 14:56 |
| coreycb | danpawlik, this is the xenial version of ceph-mds: 10.2.5-0ubuntu0.16.04.1 | 14:56 |
| danpawlik | coreycb: http://packages.ubuntu.com/xenial-updates/ceph | 14:56 |
| coreycb | danpawlik, what does 'apt-cache policy ceph-mds' show? | 14:58 |
| zul | coreycb: ok with this? http://pastebin.ubuntu.com/24139226/ | 14:59 |
| zioproto | zul: that is not good | 14:59 |
| zioproto | it will make only 50 migrations | 14:59 |
| zioproto | it is not sure that all the migrations are done | 15:00 |
| zul | zioproto: do you have a suggestion then? | 15:00 |
| zioproto | I am talking about su -s /bin/sh -c 'nova-manage db online_migrations' nova | 15:00 |
| zioproto | zul: the user should run those migration by hand | 15:00 |
| zioproto | because if you have a big database | 15:00 |
| zioproto | that is a big impact | 15:00 |
| zioproto | for this reason it is only 50 at the time | 15:01 |
| zioproto | but running just 50 migrations in that script does not make any sense | 15:01 |
| zul | zioproto/coreycb/jamespage: maybe we should not be configuring databases then | 15:01 |
| zioproto | also it is the first time I see nova-manage db sync in the package | 15:02 |
| zioproto | consider that the database migration cannot be downgraded | 15:02 |
| coreycb | zul, i think the only reason we do that for sqlite is so the service will start by default | 15:02 |
| zul | coreycb: yeah thats the reason | 15:02 |
| jamespage | zul, zioproto, coreycb: I tend to agree that we should just stop trying todo any db syncs in the packaging - the sqlite is a waste of time and just delays install | 15:02 |
| zioproto | zul: if you put a check that those commands are run only in case of sqlite than it makes sense | 15:02 |
| jamespage | the packages should be checking for sqlite before running syncs | 15:03 |
| zul | zioproto: that command only runs if there is a sqlite database anyways | 15:03 |
| jamespage | otherwise its always a remote database, and you don't want N instances of nova-cloud-controller all trying todo a db sync during the same update | 15:03 |
| danpawlik | coreycb: priv | 15:04 |
| zul | http://pastebin.ubuntu.com/24139262/ | 15:04 |
| zioproto | ok sorry from the diff it was not evident that it was inside a if block for sqlite only | 15:05 |
| zioproto | for sqlite maybe is okay | 15:05 |
| zul | if we remove the sqlite stuff from the postinst then we would have to change the dep8 tests as well | 15:05 |
| coreycb | jamespage, zul: i'm ok with dropping that db migration from the package. the problem though is we probably will have services that don't start by default. | 15:05 |
| coreycb | zul, yeah you'd probably have to use sqlite in the dep8 tests | 15:05 |
| * jamespage spin out another thought | 15:05 | |
| jamespage | maybe we should not start services by default then | 15:05 |
| zul | coreycb: or we can do mysql | 15:05 |
| zul | jamespage: yes please | 15:06 |
| jamespage | zul: using mysql in the dep8 tests is a good idea (see neutron) | 15:06 |
| zul | jamespage: manila already uses mysql in the dep8 tests btw | 15:06 |
| zul | so its pretty easy to do | 15:06 |
| jamespage | zul: yeah there are a few projecst that just don't play nice with sqlite | 15:06 |
| jamespage | zul: my original plan was to re-produce the functional test env for dep-8 tests | 15:07 |
| zul | jamespage: ack | 15:07 |
| jamespage | based on the openstack gate specs | 15:07 |
| jamespage | but never quite got to that | 15:07 |
| zul | im at ltitle hestiant to change services not to start this close to the end of the cycle though | 15:08 |
| zul | but that online migration should be ok for sqlite databases though | 15:08 |
| coreycb | zul, i'd hold off until pike for that | 15:10 |
| cetex | so, i'm working on improving our automation a bit, we're currently building AMI's from scratch with a concoction of debootstrap and a couple of bash scripts. | 15:13 |
| cetex | got stuck.. | 15:36 |
| cetex | what i'm wondering is: how does ubuntu currently provision their AMI's they have in aws? | 15:37 |
| cetex | what tooling is used to build the ami? | 15:37 |
| freakynl | Hi, I just installed Ubuntu 16.04 LTS fresh and updated it as a VMware VM. Gave it a reboot now, it's been hanging for over 5 minutes now on 'A start job is running for Unattended Upgrades Shutdown'. It's just been fully updated... Is this normal? It's a bare installed, removed open-iscsi (why is that running by default anyways?) and LVM (same question, don't use LVM) and nothing else yet. | 16:22 |
| drab | any recommendation to figure out why a container takes a long time to stop? | 16:46 |
| drab | I've had an idle container take 1.5min to stop and I'd like to figure out why | 16:46 |
| drab | but couldn't see anything in lxc.log for the container or lxd.log that indicate reasons, just start stopping and stopped times | 16:47 |
| nacc | freakynl: ubuntu server or desktop? | 16:50 |
| nacc | cetex: may want to ask in #ubuntu-server | 16:50 |
| drab | I thought this was ubuntu-server :) | 16:50 |
| rbasak | It is :) | 16:51 |
| rbasak | It might not be container-specific. If there's a service that fails to stop and systemd waits to kill it, then that could cause a delay too. | 16:51 |
| Pici | nacc is lost | 16:51 |
| drab | rbasak: yeah, I thought that may be the case, and maybe it just happened that one time for non lxc related reasons, been tarting and stopping the same container a lot for testing | 16:52 |
| drab | just trying to make sure I iron out any issue before moving this to prod | 16:52 |
| nacc | Pici: just responding to older messages :) | 16:53 |
| nacc | Pici: and yes to the latter! :) | 16:53 |
| nacc | cetex: sorry about that! | 16:53 |
| drab | like they say, the best way to find oneself is to lose oneself in the service of others ;), seems a good fit for ubuntu/freenode help desk support :) | 16:54 |
| === vonsyd0w_ is now known as vonsyd0w | ||
| cetex | nacc: no problem, i just wish someone knew how they do it ;D | 17:56 |
| nacc | Odd_Bloke: --^ amazon ami generation question | 18:24 |
| mamadg | what is OversignHeader means in Opendkim , i read document many time but i dont understand it , english is not my first , could anyone explain it to me ? | 18:50 |
| nacc | mamadg: you've read the manpage for opendkim.conf? | 18:54 |
| mamadg | nacc: yes i read man 5 opendkim.conf ... | 19:01 |
| antoniobeyah | i asked this question in #ubuntu but was redirected here | 19:05 |
| antoniobeyah | hello, it looks like there was an 16.04 lts release today/yesterday but I am unable to find the release notes- are they in progress? https://uec-images.ubuntu.com/releases/16.04/release-20170307/unpacked/ | 19:05 |
| antoniobeyah | 1:03 PM that directory normally has a release_notes.txt but it is empty | 19:05 |
| nacc | rcj: --^ ? | 19:05 |
| antoniobeyah | i was redirected to http://cloud-images.ubuntu.com/xenial/20170307/unpacked/release_notes.txt, I may have just been looking in the wrong place | 19:07 |
| antoniobeyah | thanks for the time, it looks like that has what I need to know | 19:07 |
| sarnold | antoniobeyah: hrm maybe you caught it at a funny time? this loads for me now https://uec-images.ubuntu.com/releases/16.04/release-20170307/unpacked/release_notes.txt | 19:25 |
| sarnold | antoniobeyah: thanks for pointing this out :) I've never seen these files but I like the look of them. useufl. | 19:26 |
| antoniobeyah | ha, must have been the case. it was certainly empty when I looked | 19:26 |
| antoniobeyah | I'm not sure the different between uec-images and cloud-images but it seems the cloud-images one had the release_notes before the uec one. I plan on using the cloud-images link going forward. ( http://cloud-images.ubuntu.com/xenial/20170307/unpacked/release_notes.txt) | 19:27 |
| nacc | yeah i assume there is some syncing going on, and there is a delay | 19:31 |
| aernan | how do I enable directory indexing of home directories with HTTPD/Apache2 from stock configuration? | 19:48 |
| _ruben | what would be the "easiest" way to mimic an ubuntu server installation using debootstrap? | 19:48 |
| _ruben | (or other tool that allows to do an install into a directory) | 19:49 |
| Odd_Bloke | antoniobeyah: nacc: uec-images is deprecated, so cloud-images.ubuntu.com should be used in preference. :) | 19:49 |
| antoniobeyah | Odd_Bloke: ah, thanks | 19:52 |
| nacc | Odd_Bloke: thanks! :) | 20:15 |
| === ddstreet_away is now known as ddstreet | ||
| yyyyy | anyone ever ran into an issue where a folder within a mount-point seems to have / mounted within it — i.e. a “loop”? | 21:22 |
| yyyyy | i have some volume group mounted on /sub/ and when i go to /sub/x i see /sub/x/bin, /sub/x/etc /sub/x/vmlinuz (!!!). there's no symbolic links on x and nothing mounted there on fstab or /proc/mounts. | 21:23 |
| yyyyy | i'm at a loss over why this is happening. | 21:23 |
| yyyyy | ubuntu server 16.04.1 | 21:23 |
| sarnold | yyyyy: check /proc/self/mounts | 21:23 |
| yyyyy | sarnold: nothing there as well for the /sub/x entry that seems to have a loop. | 21:25 |
| yyyyy | my other intuition would be some other sysadmin (or an attacker?) created hardlinks to that location | 21:25 |
| yyyyy | insanely, i can even check /sub/x/proc and every single /proc entry is there! | 21:28 |
| yyyyy | i have never seen this before, in about twenty years of using linux 8/ | 21:29 |
| jancoow | Hi there. My nfs server is starting on the wrong interface. Its on my "wan" port instead of my "lan" port | 21:31 |
| jancoow | how can I specify the correct interface for the nfs kernel server? | 21:31 |
| yyyyy | sarnold: is there some FUSE driver that could accomplish what i'm seeing? | 21:31 |
| sarnold | yyyyy: probably, but it'd show up in the /proc/self/mounts output, no? | 21:32 |
| yyyyy | i would assume so too | 21:35 |
| yyyyy | my guess is someone made some hardlinks there. i'm going to backup running services and delete that directory. since it's a loop, it shouldn't delete the file itself as there will be another entry as count for that inode. | 21:36 |
| yyyyy | awful, awful situation. | 21:36 |
| sarnold | yyyyy: making hardlinks to directories is ot permitted | 21:37 |
| sarnold | yyyyy: which is why I suspect it's bindmounts intstead | 21:37 |
| sarnold | yyyyy: but maybe symlinks could give similarly confusing output depending upon which tool is reporting the paths | 21:37 |
| yyyyy | it's definitely not a symlink for that directory. there are other files inside that are not shared with /. | 21:38 |
| yyyyy | a reboot is not getting rid of it, also. | 21:38 |
| yyyyy | i've checked with both `ls` and `exa` for the presence of links. | 21:38 |
| === JanC_ is now known as JanC | ||
| yyyyy | and a `sudo umount /sub/x` gives a `umount: /sub/x: not mounted` error. | 21:40 |
| yyyyy | plus, the files there seem to not always match the timestamps of what's on / | 21:47 |
| sarnold | oh | 21:47 |
| sarnold | did someone perhaps run debootstrap there? | 21:47 |
| yyyyy | not that i'm aware of! | 21:48 |
| yyyyy | would that account for the files? | 21:48 |
| yyyyy | i'm assuming blasting it out is no problem, then. | 21:50 |
| yyyyy | there's 80G in the /proc that's there (!!!) | 21:50 |
| yyyyy | lost connection here. to clean up a leftover debootstrap is it enough to just delete the folder? | 22:06 |
| powersj | nacc: I created a PPA for a package for someone to test, I want to also create a debdiff to attach to the bug. Should the debdiff changelog entry have the version from the ppa, what the release version should be, or does it matter? | 22:10 |
| sarnold | yyyyy: you didn't miss anything while you were away; if it -is- a debootstrap that's probably fine to delete it. or maybe someone did an rsync -a / host:/sub/x/ ? it's really worth being -clear- you know why you're deleting it.. | 22:27 |
| nacc | powersj: for sponsorship, your debdiff should be what you want uploaded | 22:35 |
| nacc | powersj: ideally the sponsor will review your debdiff, so if hte ppa version was there, they would complain and ask you to fix, or fix themsevles upon sponsoring (after discussing with you) | 22:36 |
| powersj | nacc: ok thanks! | 22:36 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!