[00:09] I am getting errors like crazy [00:09] ata3.00: status: { DRDY ERR } [00:10] is there a log somewhere with error's in it? [00:11] damn [00:11] i just bought those [00:12] i wonder is spinrite will work [00:15] Billy21: Look at "dmesg". [00:15] And whatever "spinrite" might be. [00:16] https://www.grc.com/sr/spinrite.htm [00:17] probably time to buy a new drive [00:17] spinrite only barely worked fifteen years ago [00:18] i'd be surprised if it can do anything useful on modern systems [00:23] ahahaha [00:23] alright [00:23] ill take them back === pmatulis_ is now known as pmatulis === tyhicks` is now known as tyhicks === mybalzit1h is now known as mybalzitch === devil is now known as Guest11632 === huats is now known as Guest33875 [09:26] hello! i have a Ubuntu server with 2 NICs. one in 10.10.0.0/24 (eth0) and one in 10.10.1.0/24 (eth1). i can connect to port 80 on from respective LAN, but i cant connect to 10.10.0.43 (its IP on eth0) from the 10.10.1.0/24 LAN, and vice versa [09:26] http://pastebin.com/yajee6JJ and forwarding is off. Any ideas? [09:27] (also if i turn ip-forwarding on, no replies are sent) [09:28] no it'd be up to your router to forward those packets betwee nthe lans [09:28] s/lans/subnets/ [09:28] that does sound like the expected behaviour to me [09:30] mybalzitch: yes, the router does forward the Syn packet, and it reaches the server, but the server sends no reply, so there is nothing more to forward... [09:30] shauno: it does? shouldnt the server reply? [09:36] well, I offered that but any mention of the router doing anything about it :) [09:36] but I think if I was stuck there, I'd be using tcpdump to see where the reply is actually going. whether it's going out the correct interface, whether it's going at all, etc [09:37] shauno: yes, there is output from "tcpdump -i any port 80" in the pastebin, and it shows (like i said above) that no reply is sent [09:38] and i would expect some reply, at some interface [09:38] thats why i am stuck:) [09:40] I'm not sure if it's relevant, but I notice in 'route -n' there's no gateway for 10.10.1.0/24 ? [09:45] shouldnt there be just one default gw? and it shouldnt need to know about any gw on eth1, since the default route is to eth0 [09:48] I think it should have one for each subnet, but only one will be the default [09:48] without it, a reply leaving eth1 addressed outside that subnet has no route [09:49] which won't impact things originating at this host, as they'll use the other interface. but might affect replies because they'll leave the interface they arrived on. I think. [09:51] and i think that replies will go out on whatever interface is matched by the routes. i think:) anyhow, the annoying thing is that _no_ reply at all is sent [09:52] the setup is very simple. installed a server with a static IP. and then added an interface with dhcp in /etc/network/interfaces and did ifup [09:52] baffles me why no reply is sent [10:00] hm, looks like I might be wrong about returning on the same interface. that's all that jumped out at me, I'm afraid [10:05] Good morning. [10:08] read this: http://jensd.be/468/linux/two-network-cards-rp_filter and tried the "quick and dirty solution", which seems to work for me! :) -> sudo sysctl net.ipv4.conf.all.rp_filter=2 === unreal_ is now known as unreal === Guest11632 is now known as devil__ === RickyB98 is now known as Raspberril === Raspberril is now known as RickyB98 [14:12] hallyn, I'm getting a pass with a later xen version for libvirt on 14.04 [14:12] * jamespage ponders whether to include xen as well [14:13] * smb wonders whether he should worry about that === bladernr_ is now known as bladernr [14:58] hi [14:59] coreycb, I'd like to add dh-python to mitaka-staging for a bit to see if it backports ok - my initial ppa testing as good [14:59] i have i gma500 with xubuntu 15.10 but this card(gma) has the problem [14:59] well it backports ok - just want to ensure everything else does as well! [14:59] jamespage ok yeah I saw something needed a new version of it [15:00] python-cffi wants it and really does need it... [15:00] jamespage, sounds good [15:00] who help me with gma500? [15:06] jamespage: I think keystoneauth1 is not updated enough in b2 [15:06] we're having issues in our CI for some tests [15:06] EmilienM, ack - let me take a peek [15:06] jamespage: also, heat is still failing but I guess it's because you did not update it yet [15:07] EmilienM, fixed the first problem, hitting a different one now [15:08] EmilienM, its rebuilding atm - I'll let you know shortly [15:08] cool [15:08] jamespage: jobs are non voting for us [15:08] jamespage: except 2 scenario, that work [15:08] ack [15:17] EmilienM, you should get aodh and libvirt updates in the next few hours as well [15:17] cool [15:17] been fixing up some backport failures... [15:17] and ironic is still failing for us [15:17] I PM'ed you last night [15:18] yeah [15:18] heat is still ftbfs [15:19] jamespage: ironic, not heat [15:19] I know [15:20] but I was looking at heat first [15:20] ok thanks [15:26] EmilienM, struggling with the ironic one: [15:27] http://logs.openstack.org/65/262965/14/check/gate-puppet-openstack-integration-scenario001-tempest-dsvm-trusty/af0c0ef/logs/ironic/ironic-conductor.txt.gz#_2016-01-04_01_20_18_979 [15:27] looks like its loading the fake driver.... [15:27] and there are no errors in the ironic log.... [15:29] https://bugs.launchpad.net/cloud-archive/+bug/1530869 [15:29] Launchpad bug 1530869 in Ubuntu Cloud Archive "ironic-conductor does not load fake driver" [Undecided,New] [15:29] yes that's what I'm looking at [15:29] let me find again the logs [15:29] in conductor somewhere [15:30] http://logs.openstack.org/65/262965/14/check/gate-puppet-openstack-integration-scenario001-tempest-dsvm-trusty/af0c0ef/logs/ironic/ironic-api.txt.gz#_2016-01-04_01_29_30_062 [15:30] Client-side error: No valid host was found. Reason: No conductor service registered which supports driver fake. [15:30] I see that in the console log, but I have no idea why that happens - the backend looks to be configured with fake [16:07] zul, can you add a bug subscriber for python-sphinx-argparse? [16:08] coreycb: url? [16:09] zul, http://launchpad.net/ubuntu/+source/sphinx-argparse === eein_ is now known as eein === Monthrect is now known as Piper-Off === jdstrand_ is now known as jdstrand === Piper-Off is now known as Monthrect [17:16] jamespage: yeah, that was my hunch [17:16] jamespage: suppose you could ask smb for sure, but ... [17:16] hallyn, its ok - backported xen as well [17:16] we have precident for that [17:18] jamespage, I am not sure which version to where (guessing Xenial to Trusty cloud-archive). Only problematic twiddle might be to work around the Trusty environment to inject the wrong LD/CCFLAGS [17:19] smb, yes === jgrimm is now known as jgrimm-afk [17:49] How is zfs support for 14.04 [17:50] just add ppa and it works, in my experience [17:50] the *support* situation I'm less clear on [17:50] https://wiki.ubuntu.com/ZFS [17:51] Alright cool thanks [17:52] zzxc: I'm making a random guess that it's 'available' but like all PPAs not 'officially supported' in 14.04 [17:52] given that 15.10+ appears to have ZFS available in the repos based on that wiki page [17:52] that page was edited recently so... [17:53] Yeah. As long as it's fairly stable it should work. [17:54] which i don't think i can attest to [17:54] I've been running it on several servers for a few years [17:54] * teward still uses ext4 for everything heh [17:54] well, 14.04 + zfs is a little bit newer, but still fairly fresh [17:55] ehm, a year or so === Guest67703 is now known as me_ === me_ is now known as med_ === jgrimm-afk is now known as jgrimm === Guest71094 is now known as bc2946088 [18:24] Is there a default login created with ubuntu autopilot for the openstack dashboard? The account created during installation only allows me to administer a project, not the actual system. === Guest63721 is now known as mfisch === mfisch is now known as Guest2004 [18:26] oddly some of the servers that I deploy from a template show the motd with the ip address and other servers from the same template do not show the ip address in the motd. I have re-run /etc/update-motd.d/50-landscape-sysinfo on all the servers and the IP address is present in all cases. Oddly though, two of the machines still will not show the ip address in the motd after a reboot. Does anyone have a suggetion? === Guest2004 is now known as mfisch [18:28] update-motd [21:05] coreycb oslo.utils is ready for review & upload [21:05] ddellav, thanks I'll take a look === Isla_de_Muerte is now known as NwS [21:32] ddellav, can you rebase your changes on the current oslo.utils? also there's a 3.7.0 you may want to pick that up instead. [21:34] coreycb will do, i did that update awhile ago. [21:35] ddellav, yeah looks like a change snuck in since then [21:42] coreycb, was the plan to update to eventlet 1.18.4? [21:42] jamespage, we need to but I haven't looked into the test failures [21:43] coreycb, want me to peek? [21:43] jamespage, sure! [21:43] coreycb, working ftbfs for uca [21:43] coreycb, heat is failing in proposed [21:44] and ceilometer never event gets to staging [21:44] and libvirt is busted in staging... [21:44] * jamespage sighs [21:44] jamespage, ok I can take some of those from you if you want [21:59] hallyn, hey - do you have an adversion to me doing a libvirt upload with a fixed virtlockd init script? its busted for trusty right now and not used on xenial... [22:03] coreycb, gee - python$i setup.py test || true [22:04] jamespage, yeah that's useful! [22:06] ddellav, python-positional sync'd, we'll need an MIR for that eventually [22:06] coreycb, Servname not supported for ai_socktype [22:06] is due to missing /etc/services in the build env [22:06] zmq we'll disable anyway [22:07] jamespage, I think we hit that before [22:07] jamespage, needs netbase package as a dep? [22:07] coreycb, yes - sorting that now [22:16] coreycb, https://launchpadlibrarian.net/244270611/buildlog_ubuntu-xenial-amd64.python-eventlet_0.18.4-1ubuntu1~ubuntu16.04.1~ppa201602292210_BUILDING.txt.gz [22:16] better [22:16] coreycb, I'll upload that now === keithzg__ is now known as keithzg [22:23] jamespage: no; i can't recall whether i'd asked smb to upload something or not... [22:24] hallyn, I basically modelled it on the libvirt-bin and virtlogd init scripts - appears to work ok [22:24] jamespage: cool - thanks [22:24] hallyn, I'll sneak one in then ;-) [22:25] coreycb, that should be libvirt fixed... [22:26] coreycb, EmilienM reported some problems with ironic as well, but I'm not familiar with that at-all [22:27] coreycb, that eventlet upload should unblock some of your oslo bits [22:27] coreycb, (mr core dev ;)) [22:41] ok... since I messed up and build a RAID out of shitty drives... is there a command to break the RAID array and then salvage the data on the off the drives? [22:41] Billy21: which raid level? [22:43] the mirror one [22:43] 0 [22:43] i think [22:43] mirror is 1, stripe is 0. [22:43] yeah 1 [22:51] i know in M$ i can just see the files when the raid is broken on a drive [22:51] i think that is because of ntfs [22:52] but I am not sure if the raid is broken if the linux file structure can be read [23:05] Billy21: On a raid1, yes, you can access the disk without issues. [23:59] Heya guys, again a quick question. I've fixed logrotate and it seems to work fine, but the HDD seems to fill instead of clearing. For example there was a 1GB log file, it got compressed into 50MB but instead of giving me back 1GB I've lost 50MB :/ any ideas what's going on?