[02:36] <notuvo> #windows-server
[03:40] <edux> Hi, I had a weird issue with a 14.04, it was not properly shutdown (server lost power) and when trying to get it back boot process was stuck with errors mount /tmp, we did boot in rescue mode to run fsck but that didn't fix the issue, we had to drop to a root shell, then exit it and 'continue boo' from friendly-rescue, it completed the boot, then a normal reboot worked fine.
[03:40] <edux> I am trying to find which commands does friendly-recovery execute to clean tmp, any idea where is that?
[03:59] <arooni> what is the recommended file type for 14.04 server?  ext3 or ext4 ?
[04:01] <pmatulis> ext4
[04:06] <edux> ?
[06:47] <cpaelzer> good morning
[08:42] <munsking> Hello, has anyone here ever tried http://www.cavium.com/ThunderX_ARM_Processors.html ?
[10:00] <jamespage> rbasak: that's alot of CVE's - https://launchpad.net/ubuntu/+source/percona-xtradb-cluster-5.6/5.6.34-26.19-0ubuntu1
[10:01] <jamespage> +1 on the suggestion to handle this as a security update...
[10:04] <Thorn> hello
[10:05] <Thorn> I copied files from /boot = /dev/sda1 to /boot1 = /dev/sdb1, how do I install grub on /dev/sdb so that it boots from /dev/sdb1? (existing grub on /dev/sda should boot from /dev/sda1)
[10:07] <Thorn> in other words I want identical boot partitions on both drives so that the system can boot from either (this is not multiboot)
[10:09] <Thorn> (root is raid1 made of /dev/sda2 and /dev/sdb2)
[14:23] <frickler> zul: do you have a build of horizon 10.0.2 for Newton UCA available somewhere? I tried to build locally but failed to get the *beep* xstatic setup right
[14:23] <zul> frickler: lemme check
[14:27] <zul> frickler: https://launchpad.net/~zulcss/+archive/ubuntu/junko
[14:29] <frickler> zul: great, thx, will check when it finishes building
[14:34] <coreycb> zul, i'm working on a dep8 failure for yakkety neutron
[14:35] <coreycb> zul, ah you have a new version in the queue, maybe we should get the earlier sru rejected and update your version in the queue ( i may have a fix, not sure yet )
[14:36] <zul> coreycb: sure go ahead
[14:36] <coreycb> zul, ok
[14:36] <hhee> guys, how can i can set history from different terminals, on same server, will be written to one file?
[14:41] <ikonia> hhee: that should already happen
[14:41] <ikonia> .bash_history
[14:45] <hhee> ikonia, got it, it wrote only when terminal exit?
[14:46] <ikonia> hhee: thats a good question, I don't know off the top of my head
[14:46] <hhee> ikonia, thx anyway
[14:46] <hhee> ikonia, have i nice day!
[14:46] <ikonia> you too
[14:46] <hhee> a*
[15:03] <php7oci> Hi everybody :)
[15:04] <php7oci> There is somebody who know how i can install php7.0 and oci (for oracle database) on ubuntu 16.04 ?
[15:06] <patdk-wk> what is oci
[15:07] <php7oci> this is the library to connect to oracke with Oracle Instant Client
[15:07] <patdk-wk> oci8 it looks like, not oci
[15:07] <zul> jamespage: ping is there a reason why seccomp is disabled for qemu on ppc64el?
[15:07] <php7oci> yes oci8
[15:07] <cpaelzer> zul: it is no more in zesty
[15:08] <jamespage> zul: yes it needs >= 2.3.0 and I did not really want it in the UCA
[15:08] <patdk-wk> apt-get install php
[15:08] <patdk-wk> then follow the instructions from the manual for oci8
[15:08] <patdk-wk> perl install oci8
[15:08] <jamespage> zul: so cpaelzer and I agreed to disable it for ppc64el
[15:08] <cpaelzer> yeah for the backport the version dep jamespage mentioned
[15:08] <jamespage> (for the UCA that is)
[15:08] <zul> ok
[15:12] <php7oci> It's ok i have seen my problem i have an ora-24454 error because my /etc/hosts is wrong
[15:12] <php7oci> thanks for your help
[15:27] <geigerCounter> Hello and I'm back.
[15:27] <geigerCounter> Heh.
[15:28] <geigerCounter> Just installed roundcube from source and was using it to test the SMTP.
[15:28] <geigerCounter> SMTP is still broken.
[15:30] <geigerCounter> I'm using exim4 as my MTA
[15:30] <geigerCounter> Any suggestions?
[16:45] <geigerCounter> Hello
[16:45] <geigerCounter> What do I install to be able to create .zip formatted archives?
[16:46] <Pici> geigerCounter: 'zip'
[16:46] <geigerCounter> Hmm. Dunno why I thought it would be trickier than that.
[16:47] <geigerCounter> Hey do you know anything about configuring exim4, Pici?
[16:47] <Pici> geigerCounter: not off the top of my head, sorry.
[16:47] <geigerCounter> Hmm. Do you know anything about OpenSSH, SMTP, or authentication methods?
[16:47] <Pici> geigerCounter: Have you seen https://help.ubuntu.com/lts/serverguide/exim4.html ?
[16:48] <geigerCounter> Pici: Yes, I have. I'm having trouble post-install and config with actually being able to authenticate a use over tls.
[16:49] <geigerCounter> It errors and says that authetnication failed even when I use the correct password.
[16:53] <rbasak> powersj: it just occurred to me that we could cron the triage bug report and send it to the correct person. Just a thought for the future - I'm not taking the effort right now but thought I'd share the thought :)
[16:53] <coreycb> jamespage, can you promote all of newton-proposed (except neutron - dep8 failure, and nova-lxd - not sure) to newton-updates?
[17:06] <jamespage> coreycb: ok
[17:19] <powersj> rbasak: I like the idea
[18:15] <muffinslinger> Hi everyone, I am trying to get Ceph working on an autopilot openstack deployment however when I try to use any ceph commands I am getting "unable to find a keyring client.admin.keyring" has anyone come across this before?
[22:00] <geigerCounter> Anyone here know much about configuring exim4?
[22:18] <geigerCounter> Anybody?
[22:19] <nacc> geigerCounter: i would just ask and see if anyone can help
[22:21] <geigerCounter> nacc: Alright, well I've followed some guides to install and configure exim4 + dovecot + roundcube with a mysql database backend and tls to secure the connections. Everything is working as it should, except for sending out mail. SMTP won't authenticate correctly and I don't know why.
[23:02] <patdk-lap> why would sending out mail authenticate at all?
[23:04] <geigerCounter> patdk-lap: What do you mean?
[23:04] <patdk-lap> exactly what I said, you don't authenticate whe nyou send out email
[23:05] <geigerCounter> Yes you do?
[23:05] <patdk-lap> no, you don't
[23:05] <patdk-lap> if you want to send email to gmail servers, how do you expect to do that? you don't have an account with gmail, you just want to email a user there
[23:06] <geigerCounter> Are you talking about sending out mail on the command-line?
[23:06] <patdk-lap> does this mean you need to create an account with every single isp in the world to send them email?
[23:06] <patdk-lap> no
[23:06] <patdk-lap> that isn't sending out email
[23:06] <geigerCounter> Yeah, you have no idea what I'm doing apparently.
[23:06] <patdk-lap> that would be submitting email
[23:06] <nacc> patdk-lap: i think you and geigerCounter are talking about different things
[23:06] <sarnold> geigerCounter: can you point out the rfc that you're talking about?
[23:06] <patdk-lap> and if you really are talking about authentication, maybe you are talking about submission, and not sending out email
[23:07] <geigerCounter> patdk-lap: I'm talking about sending mail from my server which has a fqdn to any email address out there over SMTP
[23:07] <patdk-lap> ok, why would you need authentication for that?
[23:07] <geigerCounter> sarnold: I don't know the relevant rfcs for that.
[23:08] <nacc> geigerCounter: are you the smtp server? or are you submitting it to a smtp server remotely?
[23:08] <geigerCounter> nacc: I am running the smtp server on my server, yes. Exim4 configured for doing that.
[23:09] <nacc> geigerCounter: so the issue is in exim4 itself sending the e-mail out? or in submitting to exim4 from some client? how do you know it's an authentication issue?
[23:11] <geigerCounter> nacc: The latter, and I know it's an authentication issue because the error tells me so. "535 Authentication Failed"
[23:11] <patdk-lap> the latter is not sending out to other servers, it's accepting incoming mail from users (submission)
[23:11] <nacc> right, i think it's just a terminology mismatch
[23:11] <nacc> geigerCounter: so it's exim4 failing to authenticate your client's submission
[23:12] <nacc> geigerCounter: um, is your client using TLS correctly?
[23:12] <geigerCounter> nacc: I don't honestly know.
[23:12] <geigerCounter> Also brb
[23:13] <nacc> e.g., for mutt to use tls for gmail, i have to tell it to use some specific tls settings
[23:15] <geigerCounter> nacc: Right, and Thunderbird is configured, but due to my isp I can't test with Thunderbird. So I tried to test with my fresh roundcube installation, but I don't know enough about roundcube to know if it's using tls correctly. Roundcube can connect to my imap server, which is configured to use tls and reject non-tls connections.
[23:16] <nacc> but imap is for reading not sending
[23:16] <geigerCounter> But smtp plays by its own rules
[23:16] <geigerCounter> Yeah
[23:16] <nacc> i would try with something not so heavy as roundcube
[23:16] <nacc> e.g., mutt
[23:16] <geigerCounter> And I don't know that roundcube is properly configured to use tls because I don't know as much as I probably should about roundcube
[23:16] <geigerCounter> Also mutt?
[23:17] <nacc> right, too many variables tehre, imo
[23:17] <nacc> use seomthing you understand to send mail
[23:17] <nacc> and get to the point where your smtp server is accepting requests
[23:17] <nacc> and then add more cmomplicated somethings to send mail
[23:17] <nacc> don't use a complicated something to send mail to verify if your server can accept requests :)
[23:17] <sarnold> testing with the simplest possible tools is a good idea; openssl s_client for example..
[23:18] <patdk-lap> too complex, requires users to know smtp
[23:18] <patdk-lap> just test with soemthing sane, like thunderbird, android mail, ....
[23:18] <geigerCounter> sarnold: Tried that before.
[23:18] <geigerCounter> sarnold: It doesn't work.
[23:18] <geigerCounter> patdk-lap: "requires users to know smtp" - Does smtp have more than 10 commands now?
[23:18] <sarnold> geigerCounter: that's the good one to debug then :D
[23:19] <patdk-lap> yes, it always has
[23:19] <patdk-lap> you can convert smtp auth into the correct format in your head?
[23:19] <patdk-lap> I know I can't
[23:19] <geigerCounter> No.
[23:19] <geigerCounter> I don't think, anyway?
[23:19] <geigerCounter> I mean the process I test smtp with over s_client
[23:20] <geigerCounter> "EHLO localhost" "STARTTLS" "AUTH login"
[23:20] <geigerCounter> And then it usually fails at the AUTH stage.
[23:20] <geigerCounter> Lemme go ahead and try again, actually.
[23:20] <patdk-lap> the only thing that does auth login, is outlook
[23:21] <patdk-lap> everything else does auth plain
[23:21] <geigerCounter> I've tried auth plain before.
[23:21] <geigerCounter> But I couldn't get results.
[23:30] <geigerCounter> nacc: What's mutt again?
[23:30] <nacc> geigerCounter: just another MUA
[23:31] <nacc> geigerCounter: just what i locally have configured for TLS :)
[23:31] <sarnold> "just another"??? /me eyes nacc
[23:31] <nacc> sarnold: :)
[23:31] <nacc> geigerCounter: i mean, the greatest ever MUA
[23:31] <sarnold> it's the MUA with keybindings closest to elm that also supports MIME
[23:31] <sarnold> there we go
[23:31] <geigerCounter> Hmm.
[23:31]  * nacc sends sarnold some sort of gift basket in apologies
[23:32] <geigerCounter> How would I go about learning how to mentally convert my username to base64 encoding?
[23:32] <nacc> ... don't
[23:33] <geigerCounter> nacc: Why?
[23:33] <sarnold> because there's base64(1) for that :)
[23:34] <geigerCounter> Hmm...
[23:34] <nacc> geigerCounter: why would you do it by hand ever? it seems incredibly error-prone
[23:34] <nacc> write a script that does it, at least
[23:34] <geigerCounter> nacc: Boredom?
[23:34] <geigerCounter> nacc: Mental retardation?
[23:34] <nacc> geigerCounter: without offense, it seems like your time would be better spent debugging exim4 than base64
[23:34] <geigerCounter> I know.
[23:35] <geigerCounter> But patdk laid down a challenge.
[23:35] <geigerCounter> And I'm having a hard time ignoring it.
[23:35] <sarnold> :)
[23:35] <geigerCounter> :P
[23:35] <geigerCounter> In the meantime, yeah. I'll just do the sensible thing.
[23:37] <patdk-lap> most annoying is encoding base64 plus the needed nulls
[23:38] <geigerCounter> Yes.
[23:38] <geigerCounter> Isn't there a keyboard combination you can key to add the nulls?
[23:38] <geigerCounter> Typhon: Hey! Awesome name! :D
[23:59] <geigerCounter> sarnold: I just used s_client and successfully authenticated. :Y