[00:05] <roaksoax> catbus1: a managed subnet can have 2 types of ranges. 1. DHCP range (dynamic) and 2. reserved.
[00:05] <roaksoax> catbus1: machines with "auto assign" will get ip's from *outside* those two ranges
[00:05] <roaksoax> catbus1: machines that PXE boot for enlistment, commissioning or dhcp from MAAS, will get ip's from (1)
[00:05] <roaksoax> catbus1: there's no such thing as static range anymore
[00:06] <catbus1> roaksoax: and for the unmanaged subnet?
[00:07] <maas-user> hello there. i just installed maas on my ubuntu installation, however i cannot access the UI. can sb assist please?
[00:08] <roaksoax> catbus1: the unmanaged subnet basically says "this subnet is not managed by MAAS, but you can assign IP's in this subnet to specific machines"
[00:08] <roaksoax> catbus1: in other words, if set to "auto-assign" maas wont provide an IP
[00:08] <roaksoax> catbus1: *unless* there is a reserved range that says "this subnet is unmanaged by maas, but I have X range available to use"
[00:09] <roaksoax> catbus1: note that MAAS 2.x manages *subnets* vs managing *ranges* in 1.x
[00:09] <roaksoax> maas-user: http://localhost:5240/MAAS
[00:09] <roaksoax> maas-user: did you create a admin user too ?
[00:10] <maas-user> yes
[00:11] <maas-user> Unable to connect
[00:11] <maas-user> no service is listening on port 5240
[00:11] <roaksoax> maas-user: tail -f /var/log/maas/*.log
[00:13] <maas-user> http://paste.offsec.com/?af2f6b324e476dd4#lqowLxlvi59s5n7WYwLMp8DFBV3X3TF0cRhaxZ5xcQU=
[00:15] <maas-user> looks like having disabled ipv6 might cause this: https://bugs.launchpad.net/maas/+bug/1663340
[00:19] <maas-user> ok i am going to reboot to check if that solves the problem.
[00:20] <catbus1> roaksoax: another question, can I assume the fabric will show up as 1:1 mapping to the physical interfaces on the machine? "A fabric is a set of interconnected VLANs that are capable of mutual communication. A fabric is a logical grouping of unique VLANs. A default fabric ('fabric-0') is created for each detected subnet when MAAS is installed." <-- what does this mean?
[00:22] <roaksoax> catbus1: e.g
[00:22] <roaksoax> rack controller has:
[00:22] <roaksoax> eth0
[00:22] <roaksoax> eth0.10
[00:22] <roaksoax> etho.20
[00:23] <roaksoax> eth1.30
[00:23] <roaksoax> eth1.40
[00:23] <roaksoax> catbus1: that wil reflect on
[00:23] <roaksoax> fabric-0 with untagged, vlan10 and vlan20
[00:23] <roaksoax> fabric-1 with untaged, vlan30, vlan40
[00:24] <maas-user> ok looks like maas now works when enabling ipv6.
[00:26] <catbus1> roaksoax: OK. why don't we just list fabric-0 the interface name, and list vlan and subnet accordingly based on discovery.
[00:26] <catbus1> and  A default fabric ('fabric-0') is created for each detected subnet when MAAS is installed.'
[00:27] <catbus1> and 'A default fabric ('fabric-0') is created for each detected subnet when MAAS is installed.' is this correct statement? I could have subnet defined for each vlan, but only 1 fabric-# is created for each detected interface, right?
[00:28] <roaksoax> catbus1: the fabric auto-creating is a best case attempt to auto-discover things
[00:28] <roaksoax> catbus1: it is not exact science
[00:28] <roaksoax> catbus1: the goal of it is to try to discovery different fabric/vlans
[00:28] <roaksoax> catbus1: the goal of it is to try to discovery different fabric/vlans/subnets
[00:29] <roaksoax> and whether they can mutually communicate
[00:29] <roaksoax> catbus1: a fabric is 1 or multiple switches that are trunked to each other
[00:29] <roaksoax> catbus1: so in my example, you are saing that fabric-0 (switch1, switch2, switch3) has 3 vlans, untagged, vlan10, vlan20
[00:50] <catbus1> thanks for the explanation. i will play with it to get myself familiar with the concepts and terms.
[19:37] <ThiagoCMC> Hey guys, to change MaaS' server own IP, can I just update its /etc/network/interfaces file and reboot it?
[19:48] <roaksoax> ThiagoCMC: yup
[19:48] <roaksoax> ThiagoCMC: all you need
[19:48] <ThiagoCMC> Perfect!
[22:20] <catbus1> the cloud-init handler.py at the deployment phase shows failed posting event: finish or start of various modules. where should I look at to stop this from happening, it seems it's taking a long time for each try, at the end deployment failed due to curtin failures.
[22:22] <catbus1> nm. will move on to work on something else now.