| === frankban|afk is now known as frankban | ||
| mup | Bug #1750540 opened: [2.3, UI] Filters break when navigating among Machine, device, controller tabs <ui> <MAAS:New> <https://launchpad.net/bugs/1750540> | 11:01 |
|---|---|---|
| === rharper` is now known as rharper | ||
| mup | Bug #1750604 opened: MAAS cli incorrect "Success" message <canonical-bootstack> <MAAS:New> <https://launchpad.net/bugs/1750604> | 16:10 |
| mup | Bug #1750622 opened: PXE boot on HP DL380 Gen9 is forced into a one time Legacy Boot Mode <cpe-onsite> <MAAS:New> <https://launchpad.net/bugs/1750622> | 17:38 |
| xygnal | can I have one untagged vlan and forwards multiple subnets? or do i need one untagged vlan per subnet? | 17:45 |
| xygnal | its all untagged with relay | 17:45 |
| === frankban is now known as frankban|afk | ||
| xygnal | roaksoax: ping | 19:01 |
| roaksoax | xygnal: pong | 21:19 |
| xygnal | roaksoax: 1. I uploaded json dumps from websocket calls weeks ago, did you examine them? no replies in bug report. 2. when we test bare metal, do we need to use our prod database data, or would creating 500 fake nodes in CLI be sufficient? | 21:30 |
| roaksoax | xygnal: for 1. unfortunately at the moment we dont have the cycles to investigate more your use case. For the priority to be bumped, i would need to have a support ticket | 21:31 |
| roaksoax | xygnal: as for 2. doing 500 fake nodes | 21:32 |
| roaksoax | would be a good test | 21:32 |
| roaksoax | but those fake nodes would need to be commissioned | 21:32 |
| roaksoax | and have testing results | 21:32 |
| roaksoax | xygnal: the disadvantage to that, is that there's noone really doing power checking | 21:32 |
| roaksoax | which could also be a reason why it's overloaded | 21:33 |
| roaksoax | xygnal: also, I did quick quest to see if I could reproduce your issue in different circumstances | 21:33 |
| roaksoax | xygnal: 1. created a vm with 1 CPU + 2GB ram | 21:33 |
| roaksoax | 2. added 100 real machines to MAAS | 21:33 |
| roaksoax | 3. MAAS commissioned/deployed all of those machines | 21:34 |
| roaksoax | it eventually run out of memory, which prevented the rack from being able to power manage the machnes | 21:34 |
| roaksoax | but the UI was fully functional | 21:34 |
| catbus | bladernr: <3 | 23:37 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!