[01:38] anyone know a reason that BCM5709 would not work when doing an install? [03:11] I'm trying to test a zfs pool write speed and dd keeps telling me that "dd: failed to open" the path to the zpool [03:12] I can touch a file in the directory so why wouldn't dd be able to write there? [03:15] "oflag=direct" removing that fixes the issue. [03:16] Must be a zfs thing according to google. === runelind_q_ is now known as runelind_q [04:05] if I have a bond0 interface with lacp on, would I see those lacp messages being sent out with tcpdump? [13:55] Good day, I've just started playing with uvtool and cloud images. I'm wondering where the hostname is set? as it seems to reset on each boot [14:10] ok so its the datasource disk, i can see it has been pulled out into /run/cloud-init/instance-data.json, just not sure what format that disk is in ? best way to adjust it ? [14:18] nevermind i guess im just mixing up hostname vs fqdn, it strips out the domain [14:23] really should be able to set that fqdn though i think [16:51] omg the 'new' ubuntu server install is annoying (over a slow remote console) [17:08] hallyn: new as in wich version? [17:16] it's bionic [17:16] was trying to do 'manual' disk setup [17:17] near as i can tell, every time i go down a field, after 1s it times out and goes back to the top, and there are no shortcuts for 'back' (to reread bc i set it up in a shell) or 'done' [19:47] Hi guys, can I get your input on how to find the latest network outage on my ubuntu server? [19:54] pennTeller: what kind of outage? [19:55] TJ- my website was down for a few minutes a couple of days ago and I woud like to find out why [19:55] TJ- so at the moment it was either my ubunt server failed somehow or my router was blocking visitors [19:56] so I am tying to find from the logs if the network was marked as having gone down [19:56] pennTeller: well, start with the web-server logs, it may be the web server restarted. If not, then look at the kernel log, or various logs via journalctl [19:56] pennTeller: for a hard loss of ethernet link, check the kernel log ("journalctl -b 0 -k") [19:57] TJ- thanks, I agree. I just don't know what I should be "grepping" for [19:57] thanks for that suggestion [19:57] pennTeller: if you know the approx time it happened try using a timestamp to narrow it down [19:58] thank you, do you know any particular "greppable" termns? [19:58] pennTeller: you can do things like "journalctl -b 0 --since 2 days ago' --until '1 day ago' " [19:59] pennTeller: you can use exact dates/times in those expressions [19:59] thanks I will give that a show [19:59] pennTeller: is systemd-networkd used to manage the network? if so "journalctl -b 0 -u systemd-networkd" may help [20:00] I understand it's considered good practice to not leave SSH Host keys on servers you intend for use with cloning. So you delete them. [20:01] However, I mistakenly assumed that Ubuntu Server would *automatically* generate new host keys when SSHD starts on boot. It does not. You're left with a machine you cannot log into remotely. Is this intended behavior? [20:02] RHEL and CentOS generate the new host keys, automatically, the next time sshd gets restarted [20:04] I logged in from the hypervisor console and ran "dpkg-reconfigure openssh-server" and this was the resolution. However, that seems like it shouldn't be necessary. Am I mistaken? [20:05] SuperLag: that's the way it's done, from openssh-server.postinst script. I'm not aware of any boot-time detection/tooling to do that [20:09] TJ-: So if you're prepping a VM for cloning...what is your recourse? [20:09] the idea being you don't have a bunch of VMs with the same host key.. [20:10] SuperLag: cloud-init [20:11] SuperLag: see e.g. https://github.com/canonical/cloud-init/blob/master/doc/examples/cloud-config-ssh-keys.txt