[05:24] <mup> Bug #1691933 opened: Maas gui performance is slow <canonical-bootstack> <MAAS:New> <https://launchpad.net/bugs/1691933>
[05:27] <mup> Bug #1691933 changed: Maas gui performance is slow <canonical-bootstack> <MAAS:New> <https://launchpad.net/bugs/1691933>
[05:30] <mup> Bug #1691933 opened: Maas gui performance is slow <canonical-bootstack> <MAAS:New> <https://launchpad.net/bugs/1691933>
[10:42] <mup> Bug #1691995 opened: Unable to deploy 17.10 even with image imported  <MAAS:New> <https://launchpad.net/bugs/1691995>
[11:00] <sachi> Hi
[11:01] <sachi> i have a requirement of pxe booting different machines with different image, like in one Ubuntu 14.04 and in another 16.04
[11:01] <sachi> can any one help me or give som refrence which configuration file i need to change to achieve this ?
[13:34] <pmatulis> sach...
[13:48] <jujulearn_> Any compatibility issues maas 2.1.5 and juju-2.1.2 ?
[16:27] <xygnal> roaksoax: is is possible a box could try to enlist multiple times/show up as different sytems
[16:28] <xygnal> roaksoax: for exampe if it was unable to determine any IPMI power network settings (such as setting to DHCP during enlist) would it show up again and again as a 'new' enlisted node?
[16:34] <kiko> xygnal, I don't think so, because the IPMI MAC is the same
[16:34] <kiko> so we'd have no chance of mis-identifying it
[16:34] <xygnal> and you are certain that enlisting only happens once per IPMI mac?
[16:34] <kiko> I am pretty sure physical MAC addresses are unique in the system and can only be assigned to one system
[16:34] <xygnal> the power section of these enlisted nodes
[16:34] <xygnal> is empty
[16:34] <kiko> if the machine is already registered?
[16:35] <xygnal> it was not in MAAS before enlist
[16:35] <kiko> sorry, I may have been confused -- I'm just saying that if you boot a PXE boot an unregistered machine twice it will not track those as two machines being enlisted
[16:36] <kiko> roaksoax, and any dev should be able to confirm how it's keyed
[16:36] <xygnal> we discovered that the vendor did NOT set our DRACs to dhcp for the first wave, so we changed our enlist config to reset IPMI config to DHCP so that the DRAC comes up
[16:36] <xygnal> without manual on-site intervention
[16:36] <xygnal> thats working
[16:37] <xygnal> they show up with random server names, no power settings, but DHCP does get set and they come up on our own private DHCP server (for oob only)
[16:37] <xygnal> which is when we configure their static IP, other settings, etc
[16:38] <xygnal> what we noticed is,  some of the dracs we configured just went offline for no apparent reason.  We are trying to determine if the node re-enlisted and in doing so, re-set its network config again.
[16:53] <kiko> xygnal, interesting. how many machines is it?
[16:53] <kiko> xygnal, note that if the machines are enlisting with no power settings, something is going wrong -- ipmipower is supposed to auto-configure that for you
[17:05] <xygnal> kiko we have a enlist_userdata setting the forces the device to change to DHCP.  I am guesing that is cancelling any opportunity to record more.
[17:11] <kiko> xygnal, hmmm... I'm not sure that's right. I think ipmipower runs regardless. but maybe I'm wrong
[17:23] <xygnal> kiko I did a scan through the enlist and only saw one node re-enlist a second time, and that was after we deleted it manually from first enlist.  I think we are safe.
[17:24] <xygnal> most likely someone was removing enlisted nodes
[21:52] <mup> Bug #1664698 changed: No documented way to set the MAC for a bridge or bond <maas> <MAAS:Triaged> <nplan (Ubuntu):In Progress by cyphermox> <https://launchpad.net/bugs/1664698>
[22:04] <mup> Bug #1664698 opened: No documented way to set the MAC for a bridge or bond <maas> <MAAS:Triaged> <nplan (Ubuntu):In Progress by cyphermox> <https://launchpad.net/bugs/1664698>
[22:19] <mup> Bug #1664698 changed: No documented way to set the MAC for a bridge or bond <maas> <MAAS:Triaged> <nplan (Ubuntu):In Progress by cyphermox> <https://launchpad.net/bugs/1664698>
[22:44] <roaksoax> xygnal: no. the enlistment process at least will register a machine with all its MAC addresses. If the same machine tries to re-enlist it wont be able to provided the mac addresses already exist.
[22:46] <xygnal> roaksoax ty
[22:48] <xygnal> we are up and running on time, thanks to all you and your team's diligence in cracking bugs.
[22:52] <mup> Bug #1691995 changed: Unable to deploy 17.10 even with image imported  <MAAS:Invalid> <https://launchpad.net/bugs/1691995>