[08:21] <allenap> roaksoax: I doubt you're there, but you can get this later. No, he hasn't pulled the changes yet, but he did review some additional parts last night. I have a small amount of work to do to satisfy that, then he'll pull. Should be done today or by the beginning of next week at the latest.
[12:44] <roaksoax> allenap: hehe alright, either way it is not usable
[12:44] <roaksoax> yet
[12:45] <roaksoax> allenap: let me know when you ditch the celeryconfig stuff
[12:47] <roaksoax> rvba: howdy! so I figured it the ipmi stuff out
[12:47] <roaksoax> rvba: the difference is the interface being in use
[12:47] <allenap> roaksoax: Will do. It's actually requiring (in my head at least) quite a lot of change.
[12:48] <roaksoax> ipmitool uses the OpenIPMI interface, while ipmilan uses fence_ipmilan which uses ipmitool's lan interface
[12:48] <roaksoax> (besides the fact of course fence_ipmilan was created for clustering)
[12:48] <roaksoax> allenap: ehe ok, just make sure it lands upstream before FF
[12:49] <roaksoax> allenap: if not, we can just patch our source
[12:49] <allenap> roaksoax: When is FF?
[12:49] <roaksoax> rvba: Feature Freeze (/me just remembered you guys are not bound by the Ubuntu Release Schedule :) )
[12:50] <rvba> roaksoax: do we want to keep the two power methods then?
[12:51] <roaksoax> rvba: so my idea is this. (1.) 3 IPMI options, IPMI, IPMI-LAN, IPMI-LANPLUS, or (2.) 1 IPMI option, that has "Interface" option and is a combobox, or (3.) Interface is a text box,
[12:51] <roaksoax> rvba: this obviously means that choosing the interface means adding additional parameters to the command
[12:52] <rvba> roaksoax: then I would very much be in favor or 1.
[12:52] <rvba> s/or/of/
[12:53] <allenap> rvba: Tests pass.
[12:53] <rvba> allenap: thanks.
[12:53] <roaksoax> rvba: (2.) is IPMI Power Method, which as Interface Combobox with "OpenIPMI" "LAN" "LANPLUS"
[12:54] <rvba> roaksoax: I mean it would require some JS code if we wanted to have different parameters depending on the value on the value of a combo box
[12:55] <roaksoax> rvba: right, so we 1. duplicate template files (which will only differ on ~10 words - -I opne, -I lan, -I lanplus)
[12:55] <roaksoax> or we add the different parameters
[12:55] <roaksoax> some other way
[12:56] <roaksoax> rvba: cause I think it makes no sense to duplicate the template file
[12:56] <roaksoax> rvba: cause if that's so, then a textbox would be easier
[12:58] <rvba> roaksoax: no need to duplicate the templates, we could use the inheritance feature of tempita.
[12:58] <roaksoax> rvba: cool, can you give me an example of how this would be done?
[12:59] <rvba> roaksoax: sure, hold on.
[13:04] <rvba> roaksoax: so, here are your 3 templates: master (http://paste.ubuntu.com/1089798/), specific1 (http://paste.ubuntu.com/1089799/) and specific2 (http://paste.ubuntu.com/1089800/).
[13:05] <rvba> roaksoax: if you render 'specific1' you'll get: http://paste.ubuntu.com/1089802/
[13:07] <roaksoax> rvba: ok cool, I'll get it done and forward it your way for review
[13:07] <roaksoax> thanks!
[13:07] <rvba> roaksoax: cool
[13:21] <roaksoax> rvba: how is the master referenced in the inherit? I'm guessing by telling it the path?
[13:21] <rvba> roaksoax: yes
[13:22] <roaksoax> rvba: and how does tempita now the path for the ttemplates?
[13:24] <rvba> roaksoax: you can have a look at src/maasserver/preseed.py:load_preseed_template.  That's what we've done for the preseed templates which also uses 'inherit'.
[13:42] <roaksoax> rvba: so I just realized we do not need the complexion of inheritance
[13:42] <roaksoax> rvba: the only reason why we have various files is to satisfy cobbler, but since that's gonna be gone, we can just pass a parameter
[14:03] <roaksoax> rvba: src/maasserver/power_parameters.py is only a UI thing right? I couldn't set a power_interface parameter there that would not shown on the UI, but would set a default when that parameter is selected
[14:04] <rvba> roaksoax: it is a UI thing yes. (I'm otp btw.)
[14:48] <Daviey> hola!
[15:10] <roaksoax> Daviey: howdy! have you played with IPMI much?
[15:14] <Daviey> roaksoax: yep
[15:15] <roaksoax> Daviey: so doing ipmitool -H 1.1.1.1 -U a -P a power on is the same as doing ipmitool -I lan -H 1.1.1.1 <...etc>
[15:15] <roaksoax> right?
[15:15] <roaksoax> Daviey: for on/off/status commands does it make a difference using -I lan or -I lanplus?
[15:16] <Daviey> roaksoax: err, no user / pass?
[15:16] <Daviey> no.. but i tend to use lanplus for everything
[15:16] <Daviey> another more complete alias is also chassis power on.
[15:16] <roaksoax> Daviey: i'm just trying to figure out whether in MAAS there should only be 1 IPMI command, or IPMI, IPMI-LAN and IPMI-LANPLUS
[15:17] <roaksoax> IPMI = ipmitool -H 1.2.3.4 -U a -P a chassis power <on/off>.
[15:18] <roaksoax> IPMI-LAN = ipmitool -I lan -H 1.2.3.4 -U a -P a chassis power <on/off>
[15:18] <roaksoax> IPMI-LANPLUS = ipmitool -I lanplus -H 1.2.3.4 -U a -P a chassis power <on/off>
[15:18] <roaksoax> Daviey: does that make sense, or you think is not really necessary?
[15:19] <Daviey> roaksoax: i don't think it really matters tbh
[15:19] <Daviey> roaksoax: as i say, i've used lanplus for everything i think.. If that turns out to be an issue, we can change it
[15:19] <Daviey> roaksoax: TBH, freeipmi still needs more focus imo.
[15:20] <Daviey> the bmc discover tool of freeipmi is probably needed for the ephemeral environment anyway
[15:21] <roaksoax> Daviey: right, adding the command to use is easy, so we can have a IPMI (ipmitool LANPLUS Interface) , IPMI (freeipmi)
[15:21] <roaksoax> Daviey: etc etc
[15:22] <Daviey> cool.
[15:39] <Chocobo_> Hi all,  I have set up MAAS but every action I do on the web interface requires me to re-enter my user name and password.  Also some functions (adding ssh key, and deleting nodes) does not seem to work.  Has anyone seen this before?
[15:49] <Chocobo_> hmmm, I tried a "maas flush" recreated the superuser account, but the same problem exists.
[17:24] <Daviey> Chocobo_: no, never seen or heard of that before.
[17:24] <Daviey> Chocobo_: Have you had similar issues before?
[17:24] <Daviey> Sounds like you are not maintaining your http session.