[00:08] using trusty on an acer easystore h340. system load average according to top is around 1.09. Seems like high interrupt activity with continuous ksoftirqd processes. any known issues with this setup or fixes to bring down the load? is it normal or something to be concerned about? [00:28] fk_007: what is the system doing? [00:33] well thats what i'm trying to figure out exactly. it is running apache for http, proftp, mysql, and smb. Just used as a home server and access point for away from home... not sure why the load seems so high... it has been so since I put ubuntu on it, thinking that some hardware is basically being controlled by some default (generic) driver that is causing high cpu usage but not sure... really just looking for guidance in how to look [00:33] further into this issue [00:41] possibly a process in D state [00:41] that'll drive up the load data [00:42] or in Z state (zombie) [00:42] pastebin output of 'ps axf' === arooni is now known as arooni__ === arooni__ is now known as arooni === arooni is now known as arooni__ [00:46] * patdk-lap2 wants to be in s state [00:52] http://pastebin.com/khWSkKrx [01:02] fk_007: 6665 is in D state - take a look and try to see if it stays there - what does dmesg -T have to say? [01:21] http://pastebin.com/0pCsxh71 [01:38] hey thanks for looking into this btw! [03:00] Hi, so I've put the Ubuntu Server ISO on my flash drive using Rufus, but I don't get expert install in Grub, how do I tell it to go into expert mode? [03:10] Nevermind, I figured it out [17:03] so, I have an AWS ec2 ubuntu server that is suddenly asking for a password when I try and SSH to it, I'm SSH'ing with the ubuntu user [17:03] which has no password [17:03] what do I do now?! [17:03] some help ASAP would be appreciated [17:07] lucz: I think for EC2 they actually require you to use key auth [17:07] not password auth [17:08] using the key you download from EC2, and then provide it. It's probable that someone *set* a password and got into the server [17:08] teward: you can set it to password auth [17:08] I usually do a lot of lockdown on EC2s when I have them [17:08] including key auth [17:08] right so how do I get into the server now? it's just a dev system [17:08] lucz: if you don't have the password you're going to have to key auth [17:08] or start over [17:08] how do I key auth if it's asking for a pw [17:09] can I force it to us key auth? [17:09] I really don't want to have to rebuild it [17:10] you can *try* to use `ssh -o PasswordAuthentication=no user@ipaddr` [17:10] but I would look at the output of ssh -vv as well to see whether the server is actaully accepting the keys or not [17:11] worst case is you're locked out, and on EC2 I don't think there's a way to bruteforce your wan into changing the password, but don't quote me on that (I haven't used an EC2 in a while) [17:13] ok, I'l try that [17:13] thanks [17:13] ok I got in as another user, but now what do I do to figure out why this is happening [17:14] the user I have can access root [17:16] the user I'm logged in as is meant to be able to use a password to SSH, ubuntu user uses a key [17:19] teward: any ideas? I don't see anything odd in the sshd config file [17:21] oh weird, I restarted ssh and now it's all back to normal === vtapia_ is now known as vtapia [17:26] I have another question - how do I change the displayed bash prompt hostname - i.e ubuntu@server.internet.com~$ [17:26] how do I change server.internet.com to something else [17:33] lucz: change the hostname on the server. Or change it for your user prompt. [17:34] I need to change it for everyone [17:34] who logs in [17:34] lucz: Make sure that /home/ubuntu/.ssh/ has the authorized_keys file in it, and the OpenSSL version of the public key of the private key you use for key auth is in there [17:34] lucz: change the server hostname then. [17:35] the entire hostname of the server has to change, or you have to roll custom PS1 environment variables for each shell that's installed's global profiles [17:35] but note user level profiles can override [17:46] how do I change the hostname? [17:47] both hosts and hostname files have ip addresses, but the bash prompt says a FQDN [17:49] editing hostname and restarting the hostname service doesn't work [17:53] take a look at http://askubuntu.com/questions/87665/how-do-i-change-the-hostname-without-a-restart [17:54] (sorry i'm on a crap tablet right now - can't write out uthe full answers :/) [18:00] teward: thanks :) /etc/hosts currently reads ip - localhost === terje is now known as Guest14842 [18:02] teward: running sudo hostname newhostname doesn't do anything [18:02] teward: I've tried all of the things you normally do to change the hostname and none of them do anything [18:20] i think that's something i had issues with on EC2, i had to tell them what hostname to set my server up with and that was what it always had [18:20] but it's been a while since I set it up [18:20] lucz: the only other way is to edit the global environment profiles for all thte shells on the system (read: bash) to have a different PS1 [18:21] and then instead of the autopopulating hostname it's custom text you specify [18:32] teward: thanks for your help - I think I'll just try and ansible it, we have some playbooks set up for setting the hostname but for some reason when I run them I get permission denied public key error -__- it's always something! === PaganMin is now known as GitGud [18:52] anyone around? [18:54] Gorian: ask a real question [18:54] people usually don't reply to "anyone around" [18:54] !ask [18:54] Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience [18:54] lol, I ask a question, come back a week later to no response [18:55] so might as well make sure people are around to read it ;) [18:55] coming back a week later seems to be problem 1 :p [18:55] problem two is you aren't stopping back sooner lol [18:55] just ask your questoin [18:56] well, it was hyperbole - I've been sitting in this channel 24/7 for quite a while. Used to have an IRC bouncer and sit in IRC servers for MONTHS. They all say "just ask a question" then you ask your question once every couple weeks, and get no response for montsh [18:56] *months [18:57] anyway, what I asked a few days ago with no response yet: http://i.imgur.com/slVG9UL.png [18:58] (web client sucks, sorry. Doesn't do anything more than time stamps apparently) [18:59] repost your real questoin [18:59] don't make us view an img to try and guess it [18:59] that's another problem [18:59] well, if you read the image, there would be any guessing, since it was my "real" question [18:59] just saying, you talk about "just ask your question and then someone will answer" - but I asked that questions days ago, and no one answered yuet [18:59] *yet [19:00] and I've been sitting in this IRC server 24/7 since I asked [19:00] again, you're requiring us to read an image. what about those of us IRCing from phones or tablets where that's not easy :P [19:00] Gorian, you might want to reask such poll in #ubuntu-server [19:00] OerHeks: *cough* this is #ubuntu-server [19:00] oops mea culpa [19:00] :-D [19:00] @teward the point wasn't " [19:01] "hey, go read it here" [19:01] * teward goes to do something productive. [19:01] it was "Hey, you are claiming that if i just ask my question and wait, people will answer" [19:01] you also have to be here and sometimes *repeat* your question [19:01] and I'm countering with "I asked a question and waited, and no one answered, let alone remembered that I asked" [19:02] so, I'm telling you, sometimes it's worth verifying that there are real people to actually read the question before wasting my time [19:03] I'll copy it here, on the off-chance that people are actually active this time. [19:03] hey, wondering what people think as far as running a distributed filesystem on top of raid and something like ZFS [19:04] ZFS + RAID + Gluster with no file copies? ZFS + single disks in a pool + gluster multiple copies? ZFS + redundant pools + gluster with multiple copies for maxiumum redundancy? [19:05] losing space to RAID AND losing space to GlusterFS is a lot of lost space... but single pools in zfs means you lose out on ZFS level features like self-healing, and 0 copies in GlusterFS means I can't lose a node [19:05] so, just do both and keep throwing more disk space at it until I have enough? [19:12] using raid + zfs is pointless effort [19:12] O.o [19:13] the whole point of glusterfs is to loose diskspace [19:13] the whole point of ZFS is that it creates redundancy [19:13] most of the features of ZFS rely on some form of disk redunancy in your vdevs [19:13] yes, but why would you attempt to create redundency by using raid, and then use zfs ontop of it, and loose all of the ability of zfs to do so? === med_ is now known as Guest73831 [19:13] there is no need or requirement to have redundency [19:14] because you misinterpreted what I said and thought that I meant non-zfs raid + zfs? [19:14] you can use it like any other normal filesystem that doesn't have redundency [19:14] maybe you just misspoke your question [19:14] zfs + raid + gluster [19:14] I could, but then you lost a lot of the features of ZFS [19:14] zfs + hardware raid + gluster [19:14] or else why would you ever say raid? [19:14] or what do you mean by raid? [19:14] since we can't figure out what you are talking about, hard to answer [19:15] because just before, I mentioned ZFS + single disks (i.e. zfs not use redundant vdevs vs. zfs using them) [19:15] there is no difference, zfs uses vdevs [19:16] that's not true [19:16] if you make the vdevs redundent or not, matters not to zfs [19:16] it might matter to you, but not to zfs [19:16] you can setup vdevs in multiple ways, either a pool of single-disk vdevs, mirrored, striped, raidz1,2,3 [19:16] striped? [19:17] you being pedandtic about it just to be isn't helping at all [19:17] that is not possible [19:17] it's just vdevs [19:17] !goal [19:18] right. Is there anyone in this server that isn't just hanging around to be difficult? I have better things to do than argue with people who get off on purposefully misinterpreting things people say. [19:19] I'm likely the best qualified in this channel, considering I have been using zfs on very demanding systems for almost a decade now, and have used it with glusterfs too [19:20] but if you don't want to actually ask what you want, and instead keep beating around the issue [19:20] or maybe go join #zfsonlinux and try your luck there [19:21] you also sound like you have a giant ego that gets off on trying to create strawmen out of other people's question so that you can beat the strawman down. I'm not up for that. I have to go back to work and deal with idiots all week. [19:21] the point of fun home projects is to get away from that. [19:23] well, it is *free* help [22:39] Why is haproxy built without support for tproxy?