=== popey7 is now known as popey === Napsterbater is now known as Guest14115 === Napsterbater_ is now known as Napsterbater === jamiew0w_ is now known as jamiew0w [05:59] hey all, I'm trying to get an 802.3ad bond interface going in Ubuntu but when I restart networking it hangs for a long time eventually timing out https://i.imgur.com/CcBaESb.jpg [06:00] any idea what I'm doing wrong? my interface config looks like this https://i.imgur.com/0s7ASPL.jpg [07:14] Good morning [11:33] coreycb: jamespage: Hi there! Can we please get python[3]-openstacksdk 0.27.0 released in the stein UCA? :) === cpaelzer__ is now known as cpaelzer [14:24] LarsErikP: any patches in particular or bugs that need fixing? [14:31] coreycb: well it is the latest (last?) release in stable/stein and it seems to fix a rather nasty thread handling bug, where some horizon-dashboard would accumulate threads, and eventually kill the webserver [14:34] LarsErikP: ack if you have some pointers we can likely cherry-pick the patches. would you be able to also help test the fix once we get it in proposed? [14:34] I have not managed to get a clear answer to exactly wich commit that fixed it though.. [14:34] here is the changes: https://github.com/openstack/openstacksdk/compare/0.26.0...0.27.0 [14:34] I'll be happy to help testing! [14:36] The problem we're having is with octavia-dashboard. If you leave a browser at the load-balancer page, apache starts to accumulate threads, and will eventually run out of memory. I recently tested by installin openstack 0.27.0 with pip, and that fixed the issue [14:36] openstacksdk* [14:38] I've spoken with a few guys over at both #openstack-horizon and #openstack-sdks and they were aware of the issue, but they couldn't really tell exaclty what fixed it.. [14:38] But it is indeed fixed in 0.27.0, as far as I can see.. :p [14:40] LarsErikP: it's hard to move forward without a bug. the one commit that seems relevant has a non-existing story board number Story: 2005306 [14:40] https://github.com/openstack/openstacksdk/commit/41740827c4702e8fcaebdf6adf7a2d49a557ec14 [14:41] *seems* relevant but if you're installing from pip who knows what other updates are getting pulled in from other dependencies [14:42] https://paste.ubuntu.com/p/28wsrFjBqJ/ here is the outpupt from pip3 [14:43] so, no other dependecies was pulled, I guess? [14:45] I guess I don't fully understand the process here.. Why isn't the new versions "automatically" released in UCA when they are released? There certaninly must have been _some_ bugs leading to the release of 0.27.0 anyways..? [14:51] coreycb: btw, that storyboard number exists: https://storyboard.openstack.org/#!/story/2005306 [14:52] and seems to be relevant; and fixed by the commit you pasted? :) [14:52] LarsErikP: thanks, search fail I guess :) [14:52] \o/ [14:52] it's friday afterall ;p [14:53] LarsErikP: to answer your question we have 300+ dependencies for openstack so we can't just automatically release code all over the place in stable releases [14:54] LarsErikP: we do point releases on a regular cadence with thorough testing for core packages and generally cherry-pick fixes as needed for dependencies with testing as well [14:54] aha [14:54] I see :) [14:58] but since we found the bug, this can be done? :) [15:06] LarsErikP: Let's track this via https://bugs.launchpad.net/cloud-archive/+bug/1866361 - I'll take some time deal with this probably early next week. [15:06] Launchpad bug 1866361 in Ubuntu Cloud Archive stein "[SRU] Connections to Neutron are not closed properly in v0.26.0" [Undecided,Triaged] [15:28] coreycb: fantastic! thanks <3 and, I'll be happy to help testing, at least for the issue I am experiencing. Maybe the guy that reported the bug in storyboard is also hanging around in launchpad? [15:31] LarsErikP: I think between your testing + some regression testing on our end we'll be covered [15:38] okok :) [15:38] i'll be watching the lp bug :) [16:22] Hello, I want to ask, I set cron to be 5 0 * * * /home/ben/.bash_scripts/backup-db.sh, but it won;'t run on 00:05 instead it run on 07:05. is there any explanation about why It run on 07:05 instead of 00:05? Thank you [16:24] benl90: maybe your crond runs with a different timezone than you are using in your own session? [16:24] is it because I change the date using timedatectl and not restart the cron service and the rsyslog service could result on it? Thank you. [16:24] sdeziel, how to check that sir? [16:24] benl90: that sounds likely ;) [16:25] hmm.. ok. just let [16:25] 's see if it will trigger on 00:05 [16:25] sdeziel, so after setting the timedatectl, we need to restart rsyslog and cron service ? [16:26] benl90: actually, I'm not even sure that would let crond pick up the new TZ [16:26] so crond will still use UTC time zone? [16:27] benl90: after thinking a bit more, I think restarting cron will do [16:28] cron probably uses /etc/localtime [16:28] how to see the localtime? [16:28] ls -l /etc/localtime [16:28] oh ls -al, I run cat -_-. oops [16:28] But I agree - a restart is probably required [16:29] rbasak, it already set to the zoneinfo of Asia/Jakarta. Hmm.. [16:29] timedatectl might have changed it [16:29] seems it require restart, let's see. At first I think I wrote the wrong config. [16:29] I only mention it because it's worth checking if you're confused [16:29] benl90: yes, that's what your timedatectl probably did [16:30] hmm.. okay, let's see. Thank you [16:30] benl90: but presumably, when cron was started /etc/localtime pointed elsewhere (Etc/UTC) [16:52] rbasak: sshd over lxc exec is awesome. [16:55] rbasak: do we know what PHP we're shipping for Focal> [16:55] is it 7.3 or 7.4? [16:59] teward: I believe the transition to 7.4 is still in progress [16:59] smoser: thanks :) [17:09] rbasak: do we expect it to be 7.4 by release? [17:10] because if so I have to push a tweak to the nginx default config we ship (match socket version numbering) [17:11] teward: I believe so, yes [17:11] ok then I should push this now while it's fresh (doesn't add/remove any features, so shouldn't need an FFe) [17:11] teward: https://launchpad.net/ubuntu/+source/php-defaults/73ubuntu1 - when that migrates [17:11] ok [17:13] rbasak: should I push now assuming we're going to be landing 7.4 or should I wait? (I'd have to push a fix to make it 7.3 anyways even if we're not certain 7.4 will land in time.) [17:13] (it doesn't affect the default *runtime* but it affects the commented-out example config that would allow PHP to work) [17:13] we've had several bugs about complaints about the version strings not matching up so i'm being extra diligent on that atm [17:14] teward: personally I'd wait [17:14] ack [17:14] i'll prep it on paper my side and wait [17:15] (I keep watching those PHP Triage messages that go over the lists heh) [17:15] Only because it saves a bunch of question about the UX for users following the release pocket [17:15] bryce: FYI ^ [17:19] indeed. [17:19] rbasak: that said... [17:19] I have an idea for that too. [17:21] which is to ship a PHP config in /etc/nginx/snippets/ which would contain PHP configuration and then just have an `include` in the nginx site configuration that would include the php location directly. [17:21] which would allow easier enablement of PHP functions for users [17:21] we'd still have to *edit* the socket version there, but it'd be easier for people to enable PHP interop in their site definitions [18:12] hi [18:17] any active? [18:19] Rockwood: It's better to ask your question and wait for someone to answer :) [18:21] i am using webservices from VM's ubuntu 19.10 and my project folder (by sharing) in win 10 and i am attaching the folder to ubuntu [18:24] but i am facing issue --> https://paste.ubuntu.com/p/ZZVmFdM674/ and output is "File not found." on chrome [18:25] sdeziel, this is my question how solve this? [18:25] Rockwood: looks like your web server (apache?) doesn't have permissions to read from /var/www/html/projects/cyberia/public/ [18:27] sdeziel, i tried lots of time but its not accepting permissions [18:28] win 10 (projects) and ubuntu (projects) [18:28] sdeziel, i am using nginx [18:28] Rockwood: OK, then files are accessed as the 'www-data' user by default [18:29] Rockwood: you need to make sure this user can access that public dir [18:29] how can tell the win folder that is? [18:30] means how to add this? [18:31] what's the output of "ls -l /var/www/html/projects /var/www/html/projects/cyberia" ? [18:32] err, I wanted "ls -la /var/www/html/projects /var/www/html/projects/cyberia" [18:32] sdeziel, https://paste.ubuntu.com/p/b7FnFy4pDb/ [18:32] oh [18:32] one sec [18:33] sdeziel, https://paste.ubuntu.com/p/DB9BFtvSRW/ of https://paste.ubuntu.com/p/DB9BFtvSRW/ [18:33] lol [18:33] ls -la /var/www/html/projects /var/www/html/projects/cyberia [18:33] output of this command [18:34] Rockwood: is /var/www/html/projects shared from a host running VirtualBox? [18:34] yes i am using VirtualBox [18:37] OK, I never used those but I'll assume it's mounted "rw". Have you tried this: sudo chmod o+x /var/www/html/projects /var/www/html/projects/cyberia /var/www/html/projects/cyberia/public [18:37] on win C:\Users\Professional\projects\cyberia this is the path [18:39] sdeziel, https://ibb.co/Q9GS6CX [18:40] Rockwood: if you don't want to chmod dirs, you will have to tweak the mount options to have the share mounted with different permissions (or owner or group) [18:40] for that, you'd probably need to look in /etc/fstab [18:42] sdeziel, i dont ve any problem with change anything bcz this is my dev server [18:43] i applied this -->sudo chmod o+x /var/www/html/projects /var/www/html/projects/cyberia /var/www/html/projects/cyberia/public [18:43] but isn't help in solve issue [18:43] Rockwood: the thing is that only root and members of vboxsf can access your share which means nginx running as www-data won't be able to access those [18:44] okay [18:45] Rockwood: what do you get from: grep -F /var/www/html/projects /etc/fstab ? [18:45] nothing [18:45] cyberia@cyberia:~$ grep -F /var/www/html/projects /etc/fstab [18:45] cyberia@cyberia:~$ sudo grep -F /var/www/html/projects /etc/fstab [18:45] cyberia@cyberia:~$ [18:46] mount | grep projects [18:47] cyberia@cyberia:~$ mount | grep projects [18:47] projects on /var/www/html/projects type vboxsf (rw,nodev,relatime,iocharset=utf8,uid=0,gid=998,dmode=0770,fmode=0770,tag=VBoxAutomounter) [18:48] sdeziel, ^^ [18:49] Rockwood: not sure that's the proper way but found https://askubuntu.com/questions/323392/how-to-access-shared-folders-in-ubuntu-vmoracle-virtualbox-and-link-them-to-th [18:50] Rockwood: so I'd try: sudo adduser www-data vboxsf && sudo service nginx upgrade [18:51] sdeziel, https://paste.ubuntu.com/p/BQZxdf5w6V/ [18:51] Rockwood: try browsing the site now? [18:52] no [18:52] still the same error [18:53] File not found. [18:53] browser [18:53] Rockwood: are you still seeing the same errors in nginx logs? [18:53] i am chking that [18:55] https://paste.ubuntu.com/p/ckrxJtMM2y/ its changed now in logs i think [18:55] sdeziel, [18:56] that looks better (after 2020/03/06 18:50:39). Your problem seems to be with fastcgi now [18:56] or how you interface with fastcgi from your nginx config [18:59] fastcgi_pass 127.0.0.1:9000; [18:59] something like this [18:59] by ip and port [19:00] I'd check fastcgi's logs for clues [19:05] i am unable to find the fastcgi logs [19:07] sdeziel, i am got new error file there [19:07] sdeziel, i got new error file there [19:07] error.loges [19:07] why? its generated [19:08] I just bought a domain that is 6 characters long for the sole purpose of giving short URL's, so when I orally tell/write people links and email addresses, it's super long. What's the best solution you guys suggest, as far as putting on the webserver? [19:09] Goop, you want to run your own link shortener? [19:10] lordcirth, yes. For private purposes (in other words, no everyone can create their own link). [19:12] sdeziel, what should i do now? [19:12] Rockwood: what's in the logs? [19:12] fastcgi isn't any separate log file [19:14] Rockwood: that's surprising to me. What's the package that provides your fastcgi daemon? [19:14] php-fpm [19:15] nothing in /var/log/php*-fpm.log ? [19:28] yes before this i chk another thing i just https://paste.ubuntu.com/p/4jxWStzRDx/ [19:29] now i am ppasting logs [19:31] https://paste.ubuntu.com/p/hN4hpyrdxK/ http://php7.3-fpm.log/ sdeziel [19:31] https://paste.ubuntu.com/p/hN4hpyrdxK/ <--- php7.3-fpm.log sdeziel [19:34] Rockwood: since php-fpm also runs as www-data, you need to restart that service too [19:34] okay doing [19:36] sdeziel, same issue at browser [19:44] sdeziel, there? [19:45] Rockwood: yes, at work [19:45] oh sorry [19:46] Rockwood: what's in your nginx logs now? [19:46] chkig [19:46] teward, there's just two php packages left on the 7.4 transition list, which both have weird problems but I have ideas to try today. Then just need to finish clearing any remaining proposed migration issues. I'm expecting 7.3 can be removed from the archive next week, and am hoping for early in the week. If that helps with your planning. [19:47] Yep it does i’ll follow up next week :) [19:48] sdeziel, https://paste.ubuntu.com/p/SKbRqtJzq8/ [19:49] Rockwood: those log lines are truncated in your paste for some reason. [19:50] means? [19:51] Rockwood: I cannot see the full error [19:51] i for get the pastebinit for make output link [19:51] i forget the pastebinit for make output link [19:53] Rockwood: tail /path/to/log/file.log | pastebinit [19:55] Rockwood: that said, I think your problem matches this https://serverfault.com/questions/517190/nginx-1-fastcgi-sent-in-stderr-primary-script-unknown so the first answer should be looked at [20:48] My /etc/fstab is missing a filesystem mounted on /boot where does my server get the information needed for booting? [20:52] halvors: probably from systemd [20:53] halvors: ls -l /run/systemd/generator/ [20:53] halvors: check -.mount [20:55] where -.mount is for / root filesystem you probably have others [20:56] systemctl -t mount List currently mounted systemd units [21:29] kiokoman: the file -.mount seems right, but it still boots from the root of the btrfs filesystem not the subvolume @ [21:30] just so that i mention it, the install was installed on the root of the btrfs filesystem without subvolumes. [21:30] And i created @ and @hoome [21:30] home* [21:30] the original root filesystem still exists. [21:30] does it scan thru the filesystem or something? [21:30] Do i need to delete anything other than @ and @home? [22:12] With the newest 18.04 server installer iso it does not install on @ and @home subvolumes when using btrfs? [22:17] i wuldn't be surprised. you might need to snap refresh subiquity on the installer, but this may not help either and you'll need to use the alternative installer (debian-installer) [22:17] that is, i do not actually know, have not actually tried. [22:19] ok thanks, trying with the old installer now === JanC is now known as Guest12447 === JanC_ is now known as JanC [23:02] strange, so I have a bonded interface (mode 4) but do not see any LACP messages being sent out when doing a tcpdump [23:07] jge: what does /proc/net/bonding/foo look like? [23:07] actually erase that, I was dumping on bond0 when I do it on the child ifaces I see LACP messages being sent out [23:09] it says that is up @tds but Partner churn state is churned, is that expected? [23:11] the problem I have is that my upstream switch bundle interface is not coming up.. [23:11] hmm, that doesn't sound normal, checking a box here it's none on all interfaces [23:11] do you see lacp messages from the other side when tcpdumping? [23:12] yeah I see it in both directions [23:12] from the switch and from the server [23:44] Hello, is it unusual for the major kernel version to change between LTS minor point releases? I couldn't help but notice that 18.04 went from 4.15 to 5.0. [23:49] xar-: see https://wiki.ubuntu.com/Kernel/LTSEnablementStack [23:50] rbasak: great, thank you