=== frankban|afk is now known as frankban | ||
mup | Bug #1764322 opened: allow users to set "notes" for a node <MAAS:New> <https://launchpad.net/bugs/1764322> | 08:45 |
---|---|---|
mup | Bug #1764433 opened: JS test timeout <MAAS:New for ltrager> <https://launchpad.net/bugs/1764433> | 15:10 |
mup | Bug #1764433 changed: JS test timeout <MAAS:New for ltrager> <https://launchpad.net/bugs/1764433> | 15:25 |
mup | Bug #1764433 opened: JS test timeout <MAAS:New for ltrager> <https://launchpad.net/bugs/1764433> | 15:28 |
mup | Bug #1764440 opened: [2.4, UI, performance] Every browser hard refresh generates RPC calls <MAAS:Triaged> <https://launchpad.net/bugs/1764440> | 15:34 |
=== edmz_ is now known as edmz | ||
=== narindergupta is now known as canonical | ||
eduardo_ | Hi all, trying to find information as to how I could get the processor information and save it to the node during commissioning | 22:12 |
eduardo_ | Does anyone have any script examples to share please? | 22:13 |
roaksoax | eduardo_: what processor information is that ? | 22:20 |
eduardo_ | I'd like to get the processor name and number/model | 22:21 |
roaksoax | eduardo_: that information is already available in MAAS 2.4 :) | 22:22 |
roaksoax | eduardo_: but that said, you can create tags | 22:22 |
eduardo_ | Oh, so 2.4 pulls proc automatically? :) | 22:22 |
eduardo_ | I won't even waste my time then, thanks bud! | 22:23 |
roaksoax | eduardo_: https://docs.maas.io/2.3/en/nodes-tags | 22:23 |
ananke | when is maas 2.4 going to hit GA? | 22:42 |
ananke | ahh, I see the launchpad milestones | 22:44 |
ananke | looks like roughly 3 months from alpha 1 to ga | 22:44 |
roaksoax | ananke: beta2 is already being cut, it hasn't yet been annouced | 22:54 |
roaksoax | ananke: we are targetting RC1 for next week | 22:55 |
ananke | nice. I should go through the bug list and see if it addresses anything I've seen | 22:55 |
roaksoax | ananke: is there any specific issue(s) ? | 22:55 |
ananke | other than general random odd behavior, unresponsive web UI, failing operations, etc? :) I'm seeing a lot of weird things | 22:56 |
roaksoax | ananke: more specifics would be nice :) | 22:57 |
roaksoax | ananke: e.g. how many machines do you have ? | 22:57 |
ananke | for example, I just installed a new rack controller and joined it to the region controller. it has four interfaces. somehow maas decided to put two of the NICs on one fabric | 22:57 |
roaksoax | ananke: well, that's probably because they are in the same VLAN ? | 22:57 |
ananke | so even though it was a dual port card, I was hoping to have a separate fabric on each port, hence assigning each card on the OS different IP range: 10.41.1.0/24 and 10.41.2.0/24 | 22:57 |
ananke | nope, they go to two different switches. completely separate networks | 22:58 |
ananke | so, what do I do? first I try to create a new fabric, and I name it 'fabric-5', and I hope to assign one of the nics to that fabric | 22:58 |
ananke | except that fabric is nowhere to be found | 22:58 |
ananke | k, maybe I did something wrong. I try to create it again, but then it tells me it's a reserved fabric | 22:59 |
roaksoax | ananke: the only way the vlans should have ended in the same fabric if there would have been two reaosns: | 22:59 |
roaksoax | 1. the interface is like: eth0 eth0.1 eth0.X | 23:00 |
roaksoax | 2. maas has a beaconing protocol where it sends information over hte ntework to determine the physical layout | 23:00 |
ananke | they're enp3s0f0 and enp3s0f1 | 23:00 |
roaksoax | and it discovered that those two nics are on the same fabric | 23:00 |
roaksoax | but you should be able to move the subnet to a different fabric | 23:01 |
roaksoax | 2.3.2 has fixed issues there | 23:01 |
ananke | roaksoax: ohh, interesting. I have maas 2.3.0 | 23:01 |
ananke | wonder if I'm encountering a lot of issues already fixed | 23:02 |
roaksoax | ananke: you could be: https://launchpad.net/maas/+milestone/2.3.1 | 23:02 |
roaksoax | https://launchpad.net/maas/+milestone/2.3.2 | 23:02 |
ananke | how does one upgrade to 2.3.2? maas documentation simply says to apt install maas | 23:03 |
roaksoax | ananke: 2.3.2 is not yet available in the Ubuntu Archive. It will be SRU'd soon. In the meantime, you can get it for the MAAS Stable PPA: https://docs.maas.io/2.3/en/installconfig-package-install | 23:04 |
roaksoax | ananke: under "Package repositories" you can see the two stable repositories we have | 23:04 |
roaksoax | ananke: sudo apt-add-repository -yu ppa:maas/stable | 23:04 |
roaksoax | sudo apt-get update && sudo apt-get dist-upgrade | 23:04 |
ananke | roaksoax: thank you kindly. I'll do an upgrade and see how things behave afterwards | 23:05 |
roaksoax | sounds good | 23:05 |
ananke | otherwise, this product seems very well polished. I have a couple racks of unused hardware that are a great target for this, and my engineers want to find an on-prem cloud-like solution to offer to our researchers | 23:06 |
ananke | maas + <something> + juju may be a possible combination. I just need to figure out the middle layer, whether it's open stack or something else. I was hoping to find some existing blueprints | 23:07 |
ananke | hmm, apt-add-repository doesn't seem to exist on our basic ubuntu installs. interesting. requires apt-get install software-properties-common python-software-properties | 23:10 |
roaksoax | yeah | 23:11 |
ananke | I also need to figure out why ubuntu seems to have udev/kipimi issues on dell c6100 systems, but that's another story | 23:13 |
ananke | roaksoax: what are some recommended ways to manage kvm on individual nodes for use with maas? I was hoping there would be something already built into maas, but the only method I'm arriving at is to use juju | 23:14 |
mup | Bug #1764560 opened: Non admins can add a pod <pods> <MAAS:Triaged> <https://launchpad.net/bugs/1764560> | 23:23 |
mup | Bug #1764561 opened: Non admins are presented with the 'add machine/add chassis' froms <MAAS:New> <https://launchpad.net/bugs/1764561> | 23:23 |
ananke | 'Node failed to be deleted, because of the following error:' and there is no error... | 23:27 |
ananke | hah. another 'Can't delete fabric; the following interfaces are still connected: eth0 (unknown) on <unknown-node>, eth0 (unknown) on <unknown-node>, eth0 (unknown) on' <repeats another dozen times> | 23:53 |
ananke | I may have to start over again with the entire setup. it seems if nothing goes wrong, it seems to behave ok. as soon as there are any problems, it's a snowball effect and everything breaks | 23:54 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!