[00:36] <mup> Bug #1688354 opened: Race Condition in umount_partition() When Cleaning Up Loop Device <MAAS:New> <MAAS Image Builder:New> <https://launchpad.net/bugs/1688354>
[00:39] <mup> Bug #1688354 changed: Race Condition in umount_partition() When Cleaning Up Loop Device <MAAS:New> <MAAS Image Builder:New> <https://launchpad.net/bugs/1688354>
[00:42] <mup> Bug #1688354 opened: Race Condition in umount_partition() When Cleaning Up Loop Device <MAAS:New> <MAAS Image Builder:New> <https://launchpad.net/bugs/1688354>
[00:45] <mup> Bug #1688354 changed: Race Condition in umount_partition() When Cleaning Up Loop Device <MAAS:New> <MAAS Image Builder:New> <https://launchpad.net/bugs/1688354>
[00:48] <mup> Bug #1688354 opened: Race Condition in umount_partition() When Cleaning Up Loop Device <MAAS:New> <MAAS Image Builder:New> <https://launchpad.net/bugs/1688354>
=== frankban|afk is now known as frankban
[12:00] <martijn1234> hey there! I've got a question... I've looked at the MAAS 2.1 API docs and accessed the API itself but I can't figure out what format its returning... the /api/2.0/version/ endpoint returns json, but the rest seems to be returning some sort of binary format?
[12:02] <martijn1234> Or is this a stupid question? :p
[12:40] <martijn1234> apparently its python pickle format
=== mup_ is now known as mup
=== frankban is now known as frankban|afk
[13:10] <sanjay> hi
[13:12] <sanjay> i need help regarding deployment of nodes via maas
[13:13] <mup> Bug #1690144 opened: Failed talking to pod: int() argument must be a string, a bytes-like object or a number, not 'NoneType' <MAAS:New> <https://launchpad.net/bugs/1690144>
[13:43] <mup> Bug #1689603 changed: virsh pod creation fails with pre-existing machines <MAAS:New> <https://launchpad.net/bugs/1689603>
[14:00] <sanjay> hi
[14:15] <roaksoax> sanjay: please ask your question and someone who can help will reply
[14:16] <mup> Bug #1690154 opened: block-curtin-poweroff doesn't work <MAAS:Incomplete> <https://launchpad.net/bugs/1690154>
[14:21] <sanjay> yes
[14:21] <sanjay> after curtin installation its gets stucked
[14:29] <vasey_> mpontillo: i'm past that issue now, though when I try to deploy ubuntu 17.04 i boot to a gnu grub bootloader with a "Local" option that doesn't have anything present
[14:30] <vasey_> a manual installation of ubuntu works fine, but maas' deployment doesn't work
[14:31] <roaksoax> sanjay: can you be a bit mroe specific ? the machine finishes curtin install, reboots, but then iut doesn't boot into the OS ?
[14:32] <roaksoax> vasey_: that could be a grub issue, but if I follow, you deploy, the machine is marked as deployed but it doesn't boot ?
[14:32] <roaksoax> vasey_: or it doesn't boot after restart ?
[14:33] <vasey_> the status will remain "Deploying", it'll restart but won't boot into ubuntu post-installation, which appears to work
[14:33] <roaksoax> vasey_: that seems like a efi related issue
[14:34] <vasey_> it does. i've seen a message regarding "shimx64.efi" not being found, which looks like is related to secure boot, though that's disabled on my servers
[14:35] <roaksoax> vasey_: do you have the latest firmware ? I do recall someone a couple weeks ago mentioning that upgrading their frimware fixed that for them
[14:36] <roaksoax> vasey_: that said, I also have 1 machine which i have not yet looked at fuilly that I changed from legacy to EFI that doesn't boot, experiencing exactly the same as you are
[14:36] <roaksoax> vasey_: but the other amchines (which are exactly the same type_ do work fine
[14:36] <roaksoax> vasey_: but I was messing with the bios config
[14:36] <roaksoax> so i think it must be bios/firmware
[14:37]  * roaksoax tries to deploy 7.04
[14:40] <vasey_> roaksoax: thanks, i'll look into that now
[15:16] <sanjay> the machine finishes curtin install, reboots, but then iut doesn't boot into the OS ? yes
[15:16] <sanjay> I have already delpoyed one node with same environment, but others are not getting deployed
[19:16] <sanjay> Installation complete - Node disabled netboot	Thu, 11 May. 2017 15:13:09 PXE Request - installation	Thu, 11 May. 2017 14:45:58
[19:16] <sanjay> after this failed to deployed
[19:25] <roaksoax> sanjay: that means that the install process continued fine, but the machine rebooted and didn't contact MAAS to say "i'm actually deploed"
[19:25] <roaksoax> sanjay: do you have a console log
[19:25] <roaksoax> or any logs we can see to determine what may be wrong ?
[19:25] <roaksoax> vasey_: let me know if that makes any difference
[19:26] <roaksoax> vasey_: fwiw, i successfully deployed 17.04 on EFI with secure boot disabeld
[19:28] <sanjay> no logs are also coming
[19:28] <sanjay> just it shows failed deployment
[19:30] <roaksoax> sanjay: do you have console logs ?
[19:30] <roaksoax> sanjay: what about the full node event log ?
[19:33] <sanjay> i have full node event log
[19:33] <sanjay> May 11 19:13:11 vtnode-04 rsyslogd: [origin software="rsyslogd" swVersion="8.16.0" x-pid="2001" x-info="http://www.rsyslog.com"] exiting on signal 15.
[19:33] <sanjay> this is last line..after that nothing...
[19:34] <shewless> Hi. I've been using maas 2 for awhile now and I'm a big fan.  I have the need to add some new nodes and I want them to pxe off of a different interface then normal.  If I add maas to the appropriate fabric and "enable" dhcp on that vlan associated with that fabric will everything work like magic?
[19:35] <shewless> Can I have 2 interfaces responding to DHCP/PXE or only 1?
[19:36] <shewless> I think the "maas_url" in regiond and rackd is going to hose me.. because it's an IP that the new nodes won't have access to
[19:36] <shewless> can I make the maas_url a dns name instead of an ip?
[19:36] <shewless> or can I add multiple entries?
[19:41] <sanjay> Queried node's BMC - Power state queried: on	Thu, 11 May. 2017 15:38:53 Node changed status - From 'Deploying' to 'Failed deployment'	Thu, 11 May. 2017 15:38:32 Marking node failed - Machine operation 'Deploying' timed out after 60 minutes.
[19:46] <roaksoax> sanjay: the node failed to boot
[19:46] <roaksoax> sanjay: after initial disk installation was completed
[19:46] <roaksoax> sanjay: for that, can you get console logs ?
[19:46] <roaksoax> shewless: yes you can make it a dns name
[19:47] <roaksoax> shewless: you can have the rack provide dhcp on multiple vlans for sure too, but again, the region needs to be accessible for the machines
[19:47] <roaksoax> shewless: because the machines access directly to the region to gather metadata
[19:48] <shewless> roaksoax: thanks. is there anyway to make the region listen on multiple IPs?
[19:48] <roaksoax> shewless: the region can listen on multiple IP's, but maas_url on rackd.conf is the IP your machines will be told to access
[19:48] <roaksoax> shewless: so you can either use a dns name, or you can use a VIP that is reacheable for all the machines
[19:49] <shewless> roaksoax: hmm. yeah the dns name thing probably won't work because I would have to make it two different IPs depending on the server that's booting
[19:50] <shewless> maybe I could do a VIP.. kind of a weird scenario though
=== frankban|afk is now known as frankban
[19:53] <shewless> so I could install HA proxy on my maas server
[19:54] <shewless> then it could listen on both IPs and forward to localhost
[19:54] <shewless> or rather.. forward to the maas_url?
[19:54] <roaksoax> shewless: yes you could also do that
[19:55] <shewless> either way my nodes are going to try and directly hit maas_url though.. for the metadata right?
[19:56] <roaksoax> yes
[19:56] <roaksoax> shewless: they will all hit maas_url if they are under the same rack
[19:56] <shewless> roaksoax: Another option I have is to remove the "old" IP and just add the new one since all of my nodes will have an IP and fabric on the new one
[19:57] <shewless> if I change the maas_url to a new IP.. will I break my existing nodes?
[19:57] <shewless> for example will they stay "deployed" .. and if I reboot will they successfully boot from the hard drive?
[20:03] <roaksoax> shewless: yes the machines will stay deployed
[20:03] <roaksoax> shewless: i dont see why it should create other issues
[20:03] <roaksoax> it is only really used for the deployment process
[20:03] <roaksoax> (and commissioning, etc)
[20:03] <shewless> roaksoax: thanks for the info.
[20:04] <shewless> maas is awesome!
[20:04] <roaksoax> :) thank you!
[21:06] <sanjay> any suggestions
[21:06] <sanjay> i have removed and reinstalled maas to but still facing same issue
[21:10] <mup> Bug #1443519 changed: when bind reloading is locking, regiond complete blocks. <oil> <MAAS:Won't Fix> <https://launchpad.net/bugs/1443519>
[21:10] <mup> Bug #1688359 changed: Unable to power on machine on commissioning. <MAAS:Incomplete> <https://launchpad.net/bugs/1688359>
[21:12] <roaksoax> sanjay: check /etc/rackd.conf , does it have under maas_url an IP address that the machine you are deploying can contact ?
[21:12] <roaksoax> sanjay: apart from that, without a console log, there's not much I can do to see what may be wrong
[21:43] <mup> Bug #1690231 opened: MAAS auto-assigned gateway IP, while not allowing to reserve IP's in subnets. <MAAS:Triaged by mpontillo> <https://launchpad.net/bugs/1690231>
[21:55] <sanjay> Hi Roaksoax,  i have checked /etc/rackd.conf , there is http://MASSIP:5240/MAAS
[21:56] <sanjay> yes they can ping maas server, get dhcp ip
[21:56] <sanjay> Queried node's BMC - Power state queried: on	Thu, 11 May. 2017 17:39:38 TFTP Request - libutil.c32	Thu, 11 May. 2017 17:39:30 TFTP Request - libcom32.c32	Thu, 11 May. 2017 17:39:29 TFTP Request - chain.c32	Thu, 11 May. 2017 17:39:29 PXE Request - local boot	Thu, 11 May. 2017 17:39:29
[21:57] <sanjay> after this is will do Queried node's BMC - Power state queried:
[21:57] <sanjay> and then get failed deployment
=== frankban is now known as frankban|afk
[23:05] <mup> Bug #1690256 opened: [2.1.5] Commissioning fails using 14.04 due to invalid lsblk option <MAAS:New> <https://launchpad.net/bugs/1690256>