[00:00] ikonia, i noticed something, many of those emails are being sent to 'root' on my server which i don't use this as email user [00:00] I'm noticing an issue with a fresh install of Ubuntu Server 16.04 where the server will hang for several minutes during shutdown. Most recent message displayed is "[ OK ] Stopped LVM2 metadata daemon." [00:01] ice9: that should create a bounce though as undeliverable [00:07] ikonia, how can i make postfix log directly to mail.log? [00:08] ice9: depends on your config [00:08] if this is a standard ubuntu install, the logging should be setup fine / standard out of the box [00:09] ikonia, so if removed the line syslog_name= from master.cf, it whould revert to the default behavior? [00:10] removed.... [00:10] did you put it in ? [00:11] ikonia, i mean submission inet n - y - - smtpd [00:11] -o syslog_name=postfix/submission [00:11] did you add that line ? [00:12] actually i don't remember if it was there by default or i configured it for syslog [00:12] i think i did it [00:12] revert then [00:12] as postfix maybe trying to write directly to syslog [00:12] which will fail [00:13] ikonia, i removed it but i still see postfix is logging to syslog [00:13] there should be some logging to syslog by default as syslog will log mail events [00:15] ikonia, so postfix is logging fine to syslog, and syslog is logging fine to mail.log,etc; however it still complaining about log file [00:15] ice9: postfix is not logging to syslog [00:15] ice9: the syslog is picking up mail events [00:16] ikonia, ahh then i will disbable that in syslog conf and see [00:18] ikonia, i disabled all mail conf in syslog, restarted syslog, but still seeing postfix log in syslog [00:18] :D [00:19] ice9: no you're not [00:19] ice9: postfix is not logging to syslog [00:19] syslog is picking up mail events - there is a difference [00:23] ikonia, i stopped the syslog service but postfix didn't log to any file in /var/log [00:29] ikonia, if i rebooted, will the queued mail be lost? [00:35] question: when I update a BIND dynamic zone with rndc freeze and thaw, do I need to increment the serial in the zone file as well? [00:54] ikonia, i found the issue, it was related to dovecot log file, i emptied it and everything worked again === JanC is now known as Guest24786 === JanC_ is now known as JanC [09:18] how can I change the disk scheduler on a specific disk on a machine, permenantly? I can echo to the /sys/block... path but that won't persist after a reboot apparently and I don't want to change the global scheduler, is my only option a systemd OneShot to run that? === Rar9_ is now known as Rar9 [09:34] Good morning [09:35] icey: I'd say the tmpfiles.d is easier for that goal. [09:36] icey: https://www.freedesktop.org/software/systemd/man/tmpfiles.d.html [09:37] lordievader: cool, I have no idea how to use tmpfiles.d but it looks like it could solve the problem fairly elegantly [09:38] Yes, you add a config file telling systemd to write something to a file. [09:38] lordievader: looks like a systemd only thing? Technically this needs to support back to...trusty or precise [09:38] or just trusty if I take too long to make this change ;-) [09:38] Yes, I thought so. [09:40] any suggestions then if trusty is a target for this change as well? I'm considering changing the global scheduler since I should basically be owning the machine's disks [09:40] except for the OS disk [09:41] It might be possible that you can configure this through udev too. [09:44] it always comes back to udev doesn't it [09:44] thanks for the rabbit hole to fall through lordievader! [09:46] Makes sense that it comes back to udev, right ;) [09:49] always :) === moonligh- is now known as moonlight === JanC_ is now known as JanC [20:49] ...im trying to juju register to a controller on a new bare-metal machine to an existing machine with lxd-openstack on it.. it hangs and i cant figure out why [21:28] afternoon all. i've installed 14.04 on a supermicro server and the network adapters don't seem to work after install [21:29] i see them listed in lspci [21:29] they are 82575L if it makes a difference [21:36] m4xx: what kernel are you using? [21:36] not sure, what ever the default is for 14.04 64 [21:37] uname -a shows 4.4.0-59-generic [21:38] so that's the xenial (16.04) kernel packaged for 14.04, which is fairly recent [21:38] i installed from the mini.iso if that means anything [21:38] what is weird is it worked durring install [21:40] ls mod shows that e1000e is loaded but https://downloadmirror.intel.com/20927/eng/e1000.htm says that's for pcie network cards [21:42] those are often/usually on-board [21:43] don't you see the network adapter or doesn't it get an IP address? [21:43] ifconfig -a show's enp6s0 and enp7s0 [21:43] so they seem to be there [21:43] if i try to restart networking i get an error [21:43] says it fails to stop [21:45] "stop: Job failed while stopping" to be exact [21:45] in syslog i get "networking post-stop process (1366) terminated with status 100" [21:57] in /etc/network had eth0 as the primary network interface, i change it to enp6s0 and still no network [22:05] JanC: any ideas? [22:06] are both connected to the network? [22:06] just one [22:06] in that case I'd say add both, to make sure [22:06] i tried switching the ports [22:06] that didn't make a difference [22:07] still want me to dig out another wire? [22:07] and make sure it's configured correctly (fixed IP or DHCP, maybe routing) [22:08] it's all stock ootb [22:08] set to be dhcp [22:08] and you have DHCP on the network? [22:08] yep [22:08] using on the laptop i'm on right now [22:09] my /etc/network/interfaces: auto enp6s0iface enp6s0 inet dhcp [22:09] ifquery --list --allow auto only show's l0 [22:10] I would add enp7s0 too in there [22:11] just did, rebooting [22:11] after reboot ifconfig only show's l0 [22:11] although you should be able to see with "ip link" which one is connected [22:13] hmm [22:13] looks like i get an ipv6 address [22:13] although if ifconfig shows only 'lo' that's strange, I guess [22:14] they also don't want to auto [22:30] hmm, is there any reason why a fresh installation of xenial would be listening on ports 25, 110, 143, 993, 995? (I can't see any processes listening on those ports with netstat, but I can see them when I nmap the server) [22:35] ironhalik: nmap is not fact [22:35] ironhalik: connect to the ports - see if there really is something listening [22:35] mail - obviously yes [22:35] 993 - are you running ldap ? [22:35] 995 thats imaps (stands to reason if you're running mail) [22:35] those are all mail ports no? [22:35] 143....thats mail too [22:36] a clean install shouldn't set up things like imap services though [22:36] yeah, all mail ports [22:36] tried with ncat and could connect [22:36] ironhalik: what was listening ? [22:37] dunno, noting answered - netcat reported that the connection was successful [22:37] ironhalik: are you running an imap service ? [22:37] shouldn't be [22:38] it's a fresh install [22:38] not "shoudln't be" are you [22:38] that seems unlikley [22:38] how did you do the install ? [22:38] did you use profiles, scripts etc [22:38] the standard installer, selected base system + openssh server at package selection [22:39] ok, so I nmapped other servers [22:39] and all are showing the same [22:39] there is no way on this planet a standard install should be running imaps services [22:39] that was exactly my thought :) [22:40] check what imap service is listening [22:40] you're either scanning the wrong server, mixing up nmap's resporting of 'open' and 'stealth' or using an hosting providers 'template' installation of ubuntu with these services pre-installed. [22:40] thats the thing, nothing shows up on netstat or lsof [22:41] ironhalik: what is the netstat command you are running [22:41] netstat -tulpn [22:41] tomreyn: he said he can "connect" [22:41] sec [22:41] ironhalik: and what are you expecting to see with that ? [22:42] it's a different system then, NAT or transparent proxy. [22:43] Ok, so heres some data from my local server [22:43] ie, home server running 16.04, it's not fresh but should be running amy mail services either [22:43] http://pastebin.com/raw/h2zv9s7Y [22:43] not fresh ? [22:43] shouldn't* [22:43] you just said a minute ago it was a clean install [22:43] Im scanning multiple servers [22:44] ironhalik: show me the output of the scan in a pastebin please [22:44] http://pastebin.com/raw/h2zv9s7Y <- [22:44] ironhalik: are you doing this on the same server you're scanning [22:45] localhost [22:45] as it's connecting to/from 127.0.0.1 [22:45] what the [22:47] hmm, just ran the nmap from one of the servers against another server and the output is as expected [22:48] ssh and http open [22:48] and the rest ? [22:48] they're not showing [22:48] can nmap somehow show ports on the 'client' machine, as in the one Im scanning from? [22:49] you need to look at your setup/what you're doing as imap servers are not installed by default [22:49] ironhalik: nmap will scan what you tell it to [22:49] however based on your pastebin, it looks flawed as you're referencing localhost [22:50] yeah, something is not right [22:51] ok, so a quick recap [22:51] home mac -> xenial server scan shows mail ports [22:51] xenial server -> another xenial server shows everything in order [22:52] home mac via backup gsm connection -> xenial server shows a crapton of open ports :) [22:53] looks like something is wrong with macOS nmap port?