/srv/irclogs.ubuntu.com/2013/04/02/#maas.txt

=== bigjools_ is now known as bigjools
=== racedo` is now known as racedo
AskUbuntu_relation error between nova-compute and nova cloud controller | http://askubuntu.com/q/27713913:44
roaksoaxrvba: howdy!15:30
roaksoaxallenap: howdy! were you able to make a branch available with the changes needed for the cluster_url thing for the FPI?15:30
roaksoaxrvba: were you able to take a look at https://code.launchpad.net/~andreserl/maas/ipmi_usercreation_ilo_versions_trunk/+merge/148579 ?15:32
roaksoaxallenap: and does this make sense to you? http://paste.ubuntu.com/5670766/15:34
rvbaroaksoax: hi.  Sorry, not yet;  I know I'm a bit late with this I'll take a look tomorrow morning first thing.15:34
roaksoaxrvba: awesome! thank you!15:34
rvbaroaksoax: allenap is off today and tomorrow btw.15:35
roaksoaxrvba: plop :) thanks!!15:36
roaksoaxrvba: how was australia btw? lots of gofun?15:36
rvbaroaksoax: it was great :).  We got started on a library written in Go indeed.15:37
roaksoaxrvba: awesome!15:37
rvbaroaksoax: do you have a list of the things we need to do (for the FPI, etc) in MAAS?15:39
rvbaThere is https://code.launchpad.net/~andreserl/maas/ipmi_usercreation_ilo_versions_trunk/+merge/14857915:39
rvbaThen https://code.launchpad.net/~allenap/maas/preseed-cluster-host/+merge/149954 … ?15:39
rvbaAnd then there is bug #1148016.15:40
ubot5bug 1148016 in MAAS "Node always belongs to the same nodegroup/cluster" [Undecided,New] https://launchpad.net/bugs/114801615:40
rvbaroaksoax: These are the 3 things I have on my TODO list… anything else?15:40
roaksoaxrvba: allenap branch above as bigjools mentions it doesn';y fix the issue really, but it is an improvement to the code. I already made changes to that to support the cluster_host correctly15:41
roaksoaxrvba: so bug #1148016 is the real blocker right now15:41
ubot5bug 1148016 in MAAS "Node always belongs to the same nodegroup/cluster" [Undecided,New] https://launchpad.net/bugs/114801615:41
rvbaroaksoax: ok… allenap is working on this one IIRC.15:41
roaksoaxrvba: the bug is happening specially when the cluster controller is not managing any interfaces... hence maas doesn't know how to tell that a node belongs to a particular cluster controller15:42
roaksoaxand I tested that in the same network15:42
rvbaAll right.15:43
rvbaI think we've identified where the problem is.  The only trick is that we will need to tweak the lab's network setup after the fix for this lands.15:43
rvbaThat's what alexlist was investigating.15:43
roaksoaxbiut i'm guessing its the same case on cluster with different networks (not managing dhcp/dns)15:44
rvbaerr, allenap15:44
roaksoaxrvba: yeah he did say he was gonna make a branch available for me to test15:45
rvbaroaksoax: I can probably do that for you since he is off today… would a branch based on trunk work for you?15:46
roaksoaxrvba: yeah that would be perfect15:47
rvbaroaksoax: lp:~rvb/maas/all-interfaces15:56
rvbaroaksoax: that branch changes the way we guess to which nodegroup a node should be attached.15:56
rvbaAll the interfaces are considered, not just the managed ones.15:56
rvbaSo, if you configure the interfaces correctly (and still leaves them "unmanaged"), you should be able to get the detection working, even when using an external DHCP server.15:57
roaksoaxrvba: ok awesome! I'm gonna test it15:58
roaksoaxthanks for making it available15:59
rvbanp15:59
=== matsubara is now known as matsubara-lunch
=== matsubara-lunch is now known as matsubara
=== Catbuntu is now known as Guest23171
=== HelenCrowley is now known as Catbutnu
=== Catbutnu is now known as Catbuntu
=== kentb_ is now known as kentb

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!