[00:05] ayuda porfavor, ya instale maas en un server y tengo un nodo privicional pero no lo veo en dashboard [00:23] Bug #1552442 changed: [2.0a1] After upgrade, the Static range shows as reserved range [00:39] Question: what is the highest version of MAAS suppored on Precise? [00:41] wililupy: maas on precise is very, very, very, old [00:41] wililupy: you should use at least 1.7+ [00:42] roaksoax: ok. I was hoping atleast 1.8 would work. I was going to try to use the ppa:maas/stable [00:42] But don't want to go too far down the rabbit hole if it isn't going to work. [00:43] wililupy: ha! no way [00:43] wililupy: running precise 1.8 won't work at all [00:43] wililupy: precise doesn't even have the required dependencies to run 1.5 [00:44] dang it. Ok. thanks roaksoax. === bogdanteleaga_ is now known as bogdanteleaga [02:41] Bug #1552456 changed: [2.0a] Enabling dhcpd in two vlans on different fabrics results on dhcpd being configured only for one === mimizone_ is now known as mimizone === mup_ is now known as mup === Chr1st1an_ is now known as Chr1st1an === marlinc_ is now known as marlinc === Satyr1con is now known as Satyricon === bradm_ is now known as bradm [08:26] Hello, i try to integrated the MAAS with VMWare, my virtual machine got PXE boot but the maas didn't find any storage, it's something common? [09:28] i found a bug in the autopilot thing === dooferlad_ is now known as dooferlad [10:19] hmmmm the bind server maas runs didn't seem to update dns entries [10:19] and that broke the openstack insatll [10:19] i've corrected the zone file manually [10:19] time to try again [10:29] hmmm [10:30] so when in maas you have a machine, and maas is doing dns btw, on that machine you have 3 interfaces, 2 bonded, one not, you set the bonded one to be a static ip in your subnet/fabric/range [10:31] and the other interface is left unconfigured but is left to do pxe booting [10:31] maas dns screws up and points the dns entry for that machine to the ip of the interface used to pxe boot, but once the machine is deployed with the desirered network config the dns server doesn't update [11:09] Bug #1552658 opened: MAAS DNS fails to update with correct information [12:57] imranh_: that's expected. MAAS DNS will only provide a record for the PXE interfaces, and since you are leaving it unconfigured, it won't provide you with any [13:06] Bug #1552658 changed: MAAS DNS fails to update with correct information [13:12] roaksoax: that's a bit pants [13:16] imranh_: MAAS 2.0 provides better support for it though. It will provide an A record for the PXE interface and PTR records for the rest [13:19] hi, I'm trying to run OpenStack Liberty and MaaS on the same server, but it seems that MaaS requires python-django < 1.7, and OpenStack requires python-django >= 1.7. Is there any work around to this problem? [13:19] roaksoax: that won't fix the problem i'm having though, as the A record will still point to the unconfigured interface [13:20] imranh_: you should be able to add more A records [13:20] manually in MAAS 2.0 [13:21] roaksoax: ok [13:43] so after the machines have been commissioned and the network has been set, doing 'chattr +i' on the zone file with the correct IPs seems to be working in terms of openstack deployment, currently 66% done === JZTech101 is now known as Guest11856 === Guest11856 is now known as JZTech101 [14:30] Bug #1552719 opened: division by zero calculating usage_precentage [14:31] Bug #1552726 opened: [2.0] get_dhclient_info can raise a ProcessLookupError which makes registration fail [14:31] Bug #1552727 opened: [2.0] Hosts in dhcpd.conf is incorrect and repeats === icey_ is now known as icey [15:19] Hello there. How can i backup (and restore) the MaaS database?? [15:53] Hello. I'm having trouble navigating through the MAAS docs and I'm wondering if someone has experience on setting a custom storage layout when deploying nodes? [15:56] I tried it via the web ui and the cli but I'm not getting very far. [16:19] Bug #1552775 opened: [2.0] Lease notifier is not getting notification from dhcpd [16:31] Bug #1552782 opened: [2.0] Changing gateway_ip on subnet does not cause dhcpd.conf to be regenerated [16:31] Bug #1552789 opened: [2.0] Setting secondary_rack does not trigger the rack controller to regenerate dhcpd.conf [16:37] Bug #1552782 changed: [2.0] Changing gateway_ip on subnet does not cause dhcpd.conf to be regenerated [16:37] Bug #1552789 changed: [2.0] Setting secondary_rack does not trigger the rack controller to regenerate dhcpd.conf [16:49] Bug #1552782 opened: [2.0] Changing gateway_ip on subnet does not cause dhcpd.conf to be regenerated [16:49] Bug #1552789 opened: [2.0] Setting secondary_rack does not trigger the primary_rack to regenerate dhcpd.conf [18:48] hows it going everyone? [18:48] quick question ... possibly I'm just overlooking something here, I figured I would ask anyway .... [18:49] Is there a way to configure a node's interface mtu via the context of MAAS? [18:49] core, dev, andreserl, blake_r:^ [18:53] bdx: there is, but in 1.9+ [18:53] bdx: you can doit via the api [18:54] roaksoax: in 1.9.0+bzr4533-0ubuntu1~trusty1 ? [18:54] bdx: yes, either via the 'interface' api or the 'vlan' [18:54] bdx: iirc, if you set the vlan it will apply to all the nodes on that vlan [18:54] bdx: or you can set it per node's interface [18:56] roaksoax: nice! what is the cli command to do this? Is there docs on the cli/api usage that might entail this config? [19:02] bdx: https://maas.ubuntu.com/docs1.9/api.html#interface -> you cal look there [19:02] I can't remember the commandof the top of my head [19:03] but maas interface --help :) [19:03] roaksoax: was just scoping it. thanks! [19:07] roaksoax, I can only find how to set the mtu on a nodes interface, is there somewhere else I should be looking to set it on the vlan? [19:07] even here: https://maas.ubuntu.com/docs1.9/api.html#vlan [19:07] it seems the values are being set on a per node basis ...? [19:08] all of the api requests go to /api/1.0/nodes/{system_id} [19:12] mpontillo: ^^ [19:12] mpontillo: api to set mtu on the vlan ? [19:14] roaksoax: bdx: yeah, there is an MTU field on the VLAN; you should be able to change it via the API with a CLI command like "vlan update mtu=" [19:19] mpontillo: awesome! Am I just missing it in the api docs then? [19:21] bdx: apparently so, although I just updated the docs the other day so it is strange that they are not there [19:29] roaksoax, mpontillo: it appears I cannot update the default vlan for a fabric, .... --> http://paste.ubuntu.com/15275802/ [19:30] roaksoax, mpontillo: have you any suggestions for how I might modify mtu for default vlan? [19:31] bdx: ah yes, there's an open bug that's targetted to 1.9.2 to fix that [19:31] mpontillo: ^^ [19:32] jesus, f***. [19:32] ok [19:33] roaksoax, mpontillo: thanks for your help [19:33] bdx: that's a bummer; sorry. one possible workaround would be to create a non-default interface on the fabric and move the subnet there [19:35] mpontillo: entirely ... I have a stack deployed with a lot of services running on it supporting my organization ... I'm experiencing neutron mtu grief ... and was going to try and live migrate all instnaces, and then down each of my nodes and recommission to get the mtu set to jumbo ... [19:36] bdx: do I understand correctly that you want ALL of your VLANs to suddenly become jumbo VLANs? [19:36] from the sound of it, I could probably preform these^ ops without service interuption though [19:36] mpontillo: yea, [19:37] I've already modified all of the surrounding switches and infra to support jumbo [19:38] bdx: ok, I totally didn't just tell you to do this, but I'll leave this here https://paste.ubuntu.com/15275850/ [19:38] but I need to get my metal to 9000 w/o manualy rendering all of the nodes interfaces [19:38] mpontillo: yes! [19:38] thank you!!!!! [19:39] bdx: good luck ;-) [19:39] did the trick! -> http://paste.ubuntu.com/15275861/ [19:39] thanks again! [19:53] Bug #1552882 opened: 1.9.1: For system connected to JBOD disks, failed deployment with cannot find boot sector [20:26] Bug #1552892 opened: Default architecture is blank in UI [21:39] i am hooking up some supermicros to maas and their ipmi settings are not being detected. search the web hasn't turned up anything useful. is there a way to the the enlistment log for a node so i can see the output of the enlistment script? [21:40] dbainbri: connecting tto the serial console [21:45] @roaksoax: thx. any reason it shouldn't work (ipmi detection)? [21:48] dbainbri: one of two things. 1. no internet connectivity [preventing MAAS from installing the required tools to discover IPMI. 2. error in the discovery bug or in the BMC [21:51] roaksoax: i like my odd on 1. who needs the connectivity, maas server or the node or both? does the maas server pre-download or on demand (i.e. how do i force it to download what it needs) [21:52] dbainbri: both [21:53] dbainbri: my wild guess is that the node is trying to access the ubuntu archive, but for some reason they can't [21:53] dbainbri: have you set the upstream DNS server in the MAAS Settings page [21:53] ? [21:53] Bug #1552923 opened: API allows attaching physical, bond interface to non untagged VLAN's [21:54] dbainbri: while the machine is enlisting, try looking into maas-proxy log: /var/log/maas/proxy/access.log and see if there's traffic [21:54] roaksoax: yes to 8.8.8.8 8.8.8.4 [21:55] dbainbri: so what do you see in the proxy log when enlisting ? do you see it successfully hitting the cache and downloading the packages required? [21:56] scrolls by too quickly ;) will this be logged somewhere? [21:59] dbainbri: the enlistment output is not logged anywhere unfortunately, since maas doesn't know anything about the machine [22:00] looks like on the maas server the free ipmi packages are in /var/cache [22:00] in the proxy log i see requests for ipmi packages [22:00] dbainbri: are the requests satisfied ? [22:01] see 200 in the response, so my guess is yes [22:07] dbainbri: the next step would be to look at the serial console to see what's going on then [22:11] roaksoax: thx. will have to hook that up and try. [23:17] Trying to deploy 15.10 using maas. This(http://paste.ubuntu.com/15277124/) is my .juju/environments.yaml file. The bootstrap node comes up without an error but when I add a ubunt charm (juju deploy cs:wily/trusty) the node fails in cloud-init. I cant even ping the node. Any ideas [23:27] for ipmi discovery seeing "bad spmi legacy 0" ideas?