[03:00] mercutio: I upgraded one of my Debian VMs to Jessie last night and got systemd. It's my first time using it, and it seems fine so far [03:02] The first thing I noticed about it is that just setting "console=ttyS0" in the grub boot loader is enough for it to do everything right and start a getty on that tty, which saved me having to do it [03:02] plett: does "systemd-analyze blame" work ? [03:02] it doesn't in ubuntu, as ubuntu has only started shifting to it slowly. [03:02] that's pretty cool about the console thing [03:03] (i have ubuntu vivid on some systems, the in development ubuntu) [03:03] I had no idea that command even existed, I'll find out :) [03:05] mercutio: Yes, it appears to work. The two slowest things on my system are; [03:05] 2.252s networking.service [03:05] 1.094s exim4.service [03:05] coo. [03:05] cool even [03:06] yeah i like that [03:06] But speeding up boot times has never been that important to me, I hardly ever find myself needing to wait while watching a machine boot [03:07] it's good when something takes a really long time but yeah it doesn't usually matter for low times. [03:07] i wish servers wouldn't boot through their bios so slow [03:07] oh there's a way to show how much is bios and user space etc [03:07] oh just remove blame [03:08] Startup finished in 2.173s (firmware) + 5.558s (loader) + 1.947s (kernel) + 12.413s (userspace) = 22.092s [03:08] Startup finished in 3.435s (firmware) + 5.339s (loader) + 2.986s (kernel) + 11.376s (userspace) = 23.138s [03:08] first is hard-disk system, second is ssd system [03:08] the hard-disk system is winning hah [03:08] Startup finished in 1.449s (kernel) + 7.186s (userspace) = 8.635s [03:08] my network starts slow for some reaosn [03:09] DHCP? That's the usual reason [03:09] i wonder why it doesn't show your bios. [03:09] no dhcp [03:09] infiniband + ethernet [03:09] it waits for link state. [03:09] but it's not slow enough to bug me. [03:09] i use static ip's [03:10] Mine is a KVM guest, maybe it needs some hardware counter to measure bios and loader times which mine doesn't have [03:10] Or maybe Debian haven't built that yet :) [03:11] oh this is hardware [03:12] I've got a couple of old Asus Eee netbooks, I was going to try Debian Jessie on one of those too [03:12] i haven't got any arch linux vm's atm [04:11] Jessie has systemd now. Just a headsup. [04:11] Oh that's how this conversation started. [04:51] ugh, systemd [05:16] It's a good excuse to move more things to BSD. [05:17] Though, of course, the Ubuntu base underneath all of us has been solid. [05:19] yeah [08:19] huh, systemd-analyze is handy [08:30] It's just going to say the networking script is the slowest, every time. [08:34] mine's showing POS networkmanager by 150-200ms [08:58] Anybody else seeing some poor ipv6 HE-Level3 connectivity? Was trying to download updates and every mirror my system tried was capping at 400KB/s or so. (and since it tries ipv6 first...) [08:59] (I haven't done any sort of thorough investigation yet, just a couple of mtr's.) [09:51] Seems like it may just be me. From another location's HE tunnel, same POP, 99% identical route, it's fine. There's one L3 host in the path that differs, but I don't have a 3rd HE tunnel to rule that out. [09:54] (If anyone cares, mtr http://sprunge.us/TTPR) [10:36] brycec: that's tunnelling? [10:36] it's probably just the tunnel [10:36] How so? [10:36] like he.net tunnel? [10:36] Yes [10:36] from level 3 [10:36] the tunnel server may be getting abused or something [10:37] Except I have two tunnels on that same POP (tunnel server) [10:37] maybe load balanced [10:37] They're not [10:38] hmm [10:55] that mtr doesn't show loss [10:55] but only 10 pings [10:55] That mtr was just a -r (report) [10:56] --report-cycles=500 ? [10:56] will do 500 pings [10:57] Doubtful that it would provide anything useful. The issue is persistent, I should see constant packet loss, no need for a wide window. [10:57] even 0.5% packet loss can show as 400k/sec download speeds. [12:10] *** bitslip_ has joined #arpnetworks [12:38] appareltny openbsd has some bug with virtio devices in 5.5 and 5.6 [12:38] * mercutio hasn't hit any issues [13:33] (except with spelling ;P) [13:35] heh [13:35] it's more the typing :) [13:59] *** raptelan has quit IRC (Quit: leaving) [14:11] *** joepie91 has quit IRC (Ping timeout: 258 seconds) [14:13] *** joepie91 has joined #arpnetworks [15:07] up_the_irons: at least openbsd 5.6 has wrong timestamps on mirrors.arpnetworks.com [15:07] content looks to be the same, it's just confusing. [15:08] says 7th of august when it came out 1st november [15:08] That's what she said!! [15:08] oh wow ftp5 says the same thing [15:08] maybe it's openbsd's fault :) [15:10] other servers seem to update the parent directory time at least. [15:24] Twist: ftp5 mirrors from ARP! (j/k) [15:25] heh [15:25] they really shoudl touch the dates [15:25] even if they press cd's from older files [15:25] http://www.rhaalovely.net/up2date.html [15:26] this is kind of cool [15:26] That's what she said!! [15:26] lol @ the timezones [15:26] the timezone thing is annoying [15:26] * brycec blames the Dutch [15:26] i've hit really really outdated mirrors before [15:26] @openbsd amd64 [15:26] amd64 -> snapshots: Wed Dec 10 2014 11:59:55 GMT-0800 (PST), packages: Sat Dec 06 2014 04:43:44 GMT-0800 (PST) [15:26] ^ Pulls from ftp.openbsd.org [15:26] like a year plus :) [15:27] (well, periodically polls. The pull is not "live") [15:27] i'm using snapshot on arp [15:27] i just updated it [15:27] BryceBot announces new OpenBSD snapshots for another IRC channel. [15:27] but i was also updating another box to 5.6 [15:28] it seems faster [15:28] i think ssh or tcp or something must have changed [15:28] cos it's like a slow old server, adn it seems faster than faster servers [15:49] i think it's openssh [15:50] Perhaps it's defaulting to faster ciphers [15:51] mercutio, I've had my networking fail and say buffer full or something like that whenever I try to write out on a socket [15:52] I guess that's not a hang like the errata says though [15:52] brycec: yeh i think so, it's faster with ssh than dropbear to connect now [15:52] mkb: sounds nasty. [15:52] That's what she said!! [15:52] at least they informed about it [15:56] openbsd really needs binary updates. [15:57] err would be good with [15:57] they're classing themselves as a research OS now [15:57] mtier.org for binary updates [16:01] Specifically http://www.mtier.org/solutions/apps/openup/ [16:03] i did not know about that [16:04] And now you do, hooray! [16:04] thanks brycec [16:04] np [16:04] was it on undeadly? [16:04] I have no idea [16:04] i read undeadly a bit and occassionally check out the mailing list. [16:04] mailing list has so much volume though [16:04] mtier the company has definitely been featured [16:04] I know about it from fellow OpenBSD sysadmins and developers [16:05] ahh [16:05] it doesn't like me having http_proxy set [16:05] https_proxy isn't set [16:05] http://www.undeadly.org/cgi?action=article&sid=20110420080633 [16:05] oh that was ages ago hah [16:06] undeadly's search is broken it seems [16:06] you used google? [16:06] yes [16:06] http://undeadly.org/cgi?action=article&sid=20130509120042 [16:06] it was mentioned more recently [16:06] that's them using it as a desktop (complete with screenshots) [16:08] i used openbsd on desktop for a while [16:08] it worked pertty well [16:08] it was a while ago when video drivers were more open source friendly [16:08] I still do [16:08] we have KMS now, at least if it's not nvidia [16:09] i haven't tried openbsd on desktop in a while [16:09] maybe i should [16:09] linux's open source radeon support is really bugging me. [16:10] i can't expect openbsd to be any better, but intel onboard was much better :) [16:10] firefox is slow [16:10] I don't know if that's OpenBSD, my computer, or Firefox... [16:10] i've never seen firefox be particularly fast. [16:10] i've seen it scrolling at various speeds, but seems to have random pauses [16:10] i was using opera on openbsd [16:11] the freebsd version [16:11] i don't think that even exists anymore [16:11] Iwant Firefox because I'm able to give it all the right settings w.r.t. cookies and such. [16:12] ahh i've never worried about that [16:15] *** brycec is now known as travisCI [16:15] *** travisCI is now known as brycec [18:06] mercutio: new opera is just chromium + opera features now [18:07] i think they dropped freebsd though? [18:07] probably [18:07] and it's still closed source isn't it? [18:07] yep, that doesn't bother me in the slightest [18:07] i liked it years ago when they were the first to do a lot of stuff, ahead of other browsers [18:07] lots of cool functionality [18:10] yeah [18:10] i liked it except it kept crashing a lot [18:10] but it would show the pages before the crash when you resarted it [18:11] so i kind of just put up with it :) [18:11] That's what she said!! [18:12] mkb, m0unds: have you tried xombrero? [18:25] I have [18:26] what I like is Firefox's profiles [18:26] I have one that I normally use that has cookies, javascript, etc. off [18:26] and another that I use if I need to that has it all on, but has remove on close set to true [18:44] *** medum has quit IRC (Ping timeout: 256 seconds) [18:59] *** kevr has quit IRC (*.net *.split) [18:59] *** jbergstroem has quit IRC (*.net *.split) [19:05] *** kevr has joined #arpnetworks [19:05] *** jbergstroem has joined #arpnetworks [19:08] *** kevr has quit IRC (Max SendQ exceeded) [19:10] *** kevr has joined #arpnetworks [19:50] where does one generally store ISO images for use with xen VMs? [19:51] I think I found: /var/lib/libvirt/images [19:53] i would use /iso [19:53] or /xeniso [19:54] to my mind it's easier just having top level paths [19:57] I get the following error: https://gist.github.com/anonymous/a0f1e4e78dab4ac64d65 [19:57] Gist: "https://gist.github.com/a0f1e4e78dab4ac64d65" [19:57] my first time creating a Xen VM [19:57] eww libvirt [19:58] using virt-manager remotely from a CentOs 7 VM [19:58] host is Ubuntu [19:58] yeah i just xl [19:59] it sounds like a key issue [19:59] try sshing in with the key and adding it to the file first? [19:59] maybe it's prompting or osmething [20:00] would it make sense to install a gui on the host and create the VM locally? [20:03] xl isn't that hard [20:05] and doesn't require gui [20:05] looking into that now [20:05] libvirt i find harder to read [20:05] That's what she said!! [20:05] BryceBot: no [20:05] Oh, okay... I'm sorry. 'libvirt i find harder to read' [20:05] but can work with kvm and xen [20:06] http://xen.1045712.n5.nabble.com/xen-unstable-xl-add-some-example-configuration-files-td4956691.html [20:06] how do you view the actualdisplay? [20:06] i was finding it hard to find examples [20:06] i just use normal console [20:06] you can type xl console and it does a serial type console [20:06] like xl console virtualmachinename [20:06] what if the guest has a gui? [20:06] you can esacpe it with ctrl-] [20:06] oh gui? [20:06] i only do unix :) [20:06] on virtual machines. [20:06] like a windows VM for example [20:07] you can set it up with vnc [20:07] you do vnc=1 [20:07] and you can tell it a port number [20:07] you can also do sdl=1 for local stuff [20:07] there'll be something for that new fancy one too, but generally speaking i've just used vnc then used remote desktop once it's there [20:07] you can then either run vncviewer with remote X forwarding, or tunnel a port via ssh [20:08] or listen on a private network [20:20] this seems a bit complex for me [20:21] what part? [20:21] I like click - click gui configs [20:21] windows? :) [20:21] like virt-manager [20:21] and vmware esxi [20:21] virtualbox is good for desktop usage like that [20:22] ok well your virt-manager thing is probably about the ssh key thing [20:22] you can probably ssh forward X virt-manager from a remote server and run it locally [20:22] or you can figure out why ssh is showing that stuff and fix it [20:22] so ssh'ing in manually should let you do the accept on the key [20:24] does virt-manager have to be installed on the host even though I am using it remotely over ssh [20:26] I am actually running virt-manager on has it installed and is connecting to the host via ssh [20:26] yeah i know [20:27] that's why i said you have to ssh in manually [20:27] it's showing a ssh generic thing [20:27] which suggests you haven't accepted the remote key [20:27] as the user virt-manager ssh's in as [20:27] so it's added to host keys or whatever [20:38] ssh did not prompt to accept a key [20:39] not sure what user the virt-manager is using the ssh [20:39] should just be root [21:12] just see what ps says? [23:03] *** kevr has quit IRC (*.net *.split) [23:03] *** jbergstroem has quit IRC (*.net *.split) [23:08] *** kevr has joined #arpnetworks [23:08] *** jbergstroem has joined #arpnetworks [23:10] *** kevr has quit IRC (Max SendQ exceeded) [23:14] *** kevr has joined #arpnetworks [23:15] *** kevr has quit IRC (Changing host) [23:15] *** kevr has joined #arpnetworks