[00:24] new microcode out for intel cpus, anyone updated yet? how to :D? [00:25] https://askubuntu.com/questions/545925/how-to-update-intel-microcode-properly [00:25] this answer doesn't actually tell you how to update it [00:48] bindi: have the intel-microcode package installed and keep your system updated [00:48] you will get it via -security, I believe for your sytem [01:12] ... [01:12] that doesnt answer my question :D [01:12] well, how to update with the tool manually, is what i wanted to ask specifically [01:12] it isnt updated in the repos yet === nchambers is now known as TheMediocreTroll === TheMediocreTroll is now known as nchambers [04:14] Hello! I am trying to configure my interfaces file to get pfsense working inside KVM. Currently my /etc/network/interfaces file looks like this: https://gist.github.com/marshallford/cbf917a9cf8cbd8d23c641b04c193569 What gives? my WAN NIC passes though but the LAN isn't working at al [04:15] To follow up my question: The Ubuntu Host OS should have the ip 10.0.0.3 and allow enp8s0 (LAN NIC) to hook up to a switch or access point. I must be missing something simple... [05:58] nacc: I thought he was on it, let me check the current state of curl/http2 [06:03] nacc: no it is actually fixed by the last merge as locutusofborg and I discussed [06:03] just missing in the changelog [06:03] I'll update the bug [07:20] good mornign btw [07:20] :-) [08:17] hello [08:18] upgrading the Kernel on openstack compute nodes we had run into this https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1742630 [08:18] Launchpad bug 1742630 in linux (Ubuntu) "Booting from 4.13.0-21-generic leads to Oops: NULL pointer dereference - RIP: isci_task_abort_task+0x30/0x3e0 [isci]" [Undecided,New] [08:18] anyone else with quanta servers ? :) [08:45] Good morning [11:11] good morning === albech1 is now known as albech [14:33] gonna assume the issue with some 16.04 systems getting "bricked" with the newkernel update is fixed ? [14:41] Ussat: I thought the bricking was with newer releases, not Xenial. [14:41] There was a post-kernel-side-channel-patch issue that's been resolved, but it didn't brick. [14:41] The last kernel could be booted. (It bit my wife's desktop.) [14:43] Yeah, I haven't heard of any _bricking_ issues. [14:50] I think there was a bricking issue with new kernels on some newish systems. Don't remember the details. [14:50] Ah: https://lwn.net/Articles/741916/rss [14:51] there were problems with 4.4.0-108, which were resolved with -109. Of course, situation is still fluid-ish [14:51] I just now saw a SlashDot article talking about 16.04, so I stand corrected. [14:51] https://news.slashdot.org/story/18/01/10/1634215/meltdown-and-spectre-patches-bricking-ubuntu-1604-computers [14:51] Ah, no, that's not bricking. [14:52] That's "this kernel won't boot - boot with the last kernel" after which an update works to resolve the issue. [14:52] But yeah, that bit my wife's system. [14:54] Should perhaps never ever write an article again if you call that bricking [14:54] That occurs to me, yes. [14:54] Not that it's the first terribly worded or researched thing on slashdot :p [14:54] well, for the casual user, being unable to boot is just like bricking [14:54] Sure [14:55] But pretending to be some tech news reporting... I have more faith in 4chan [14:55] anyhow! [14:55] * hateball goes back to lurking [14:55] Just like bricking, except that it's easily fixed, as opposed to, you know, bricking. [14:57] :-) [14:58] well, I have a 17* system that upgradede fine (its a test box), just spun a 16.04 LTS and testing it now [14:58] mason, ya ya I know... [14:58] If you upgrade today, yeah, no problem. They issued errata an hour or two after my wife updated and found herself staring at a blank screen. [14:59] OK, kinda figured they would, just testing it out on my test boxes before I go to schedule prod [15:00] Sounds like the safest thing to do, yar. [15:01] Now if you and RHEL could scyn your patch releases :) [15:01] sync [15:02] You and RHEL? :P [15:02] Yes, I have both RHEL and Ubuntu systems in prod [15:02] :) [15:03] No, understood - just saying, you and RHEL. Thought you were talking to me. :P [15:03] heh , fair buff [15:03] nuff [15:03] I meant generic you [15:03] The embargo should have led to synchronization, but it was all a fustercluck. [15:03] Yar. [15:04] Xen releases fixes early, paper comes out early with details, and suddenly the embargo is off. [15:04] Some smaller projects still haven't released patches. [15:04] Slackware, the BSDs are all still lagging as far as I understand it. [15:05] Ah, I spoke too soon. DragonflyBSD has errata out now. [15:06] Hell, Vmware snuck fixes in about 2 weeks ago [15:06] Ah, didn't realize. [15:07] As for me, I want to move my infrastructure over to Raspberry Pis after this. [15:08] heh [15:08] Yea...and the best part, what else is comming now that this vector is public [15:08] Yep. [15:09] I'm having meetings with customers talking about information hygiene, data classification on multitenant systems, etc. [15:09] There's some good, traditional best practise that can help a bit. [15:09] I bet, I am in healthcare, so HIPPA etc....fun stuff [15:10] You're probably already doing everything we're recommending. [15:11] Oh ya....all sorts fun [20:18] I just got an intel-microcode update, do I need to restart my server again? [20:21] mtl: AIUI, it won't really have much impact until the next kernel drops. [21:00] mtl: you can load it using "echo 1 > /sys/devices/system/cpu/microcode/reload" rather than reboot [21:40] And check dmesg to see that it did its thing. [21:40] mtl: At the least, update-initramfs so it loads on reboots [22:31] Hello! I have a server which has been updated with the meltdown patch. I need to remove this and prevent the install in the future, any hints on how to remove this? [22:32] boxrick: Why do you need to remove it? [22:32] (You really, really don't want to run a machine without this patch. :) [22:33] This is an offline server, where performance is critical and the security is rather irrelevent [22:34] boxrick: If it's offline, how did you upgrade the kernel? ;) [22:34] I use it for number crunching, and keep it with an airgap and only use my mirror for updates when I need something specific. [22:34] Hah, quick answer. :p [22:34] I was typing that as you asked :) [22:36] boxrick: it seems like you would need to not run -updates/-security, or pin your linux-generic (or wahtever) package [22:36] boxrick: as I assume these patches will now exist forever forward [22:36] boxrick: isn't it easier to just pass nopti? [22:36] Yea good point [22:38] Cheers, will just update grub.