The belief is that Clint is going to add some event in upstart that will be fired when all 'auto' network interfaces in /etc/network/interfaces are up. Given that event named 'NEW_NETWORK_DEVICES_ARE_UP_EVENT', the patch for this bug looks something (I think) like this:
--- upstart/cloud-init-nonet.conf 2011-06-28 10:41:05 +0000
+++ upstart/cloud-init-nonet.conf 2011-07-19 17:15:51 +0000
@@ -3,7 +3,7 @@
# * to block running of cloud-init until a non 'lo' interface is up
# * timeout if one doens't come up in a reasonable amount of time
start on mounted MOUNTPOINT=/ and stopped cloud-init-local
-stop on net-device-up IFACE!=lo
+stop on NEW_NETWORK_DEVICES_ARE_UP_EVENT
task
The belief is that Clint is going to add some event in upstart that will be fired when all 'auto' network interfaces in /etc/network/ interfaces are up. Given that event named 'NEW_NETWORK_ DEVICES_ ARE_UP_ EVENT', the patch for this bug looks something (I think) like this:
--- upstart/ cloud-init- nonet.conf 2011-06-28 10:41:05 +0000 cloud-init- nonet.conf 2011-07-19 17:15:51 +0000 DEVICES_ ARE_UP_ EVENT
+++ upstart/
@@ -3,7 +3,7 @@
# * to block running of cloud-init until a non 'lo' interface is up
# * timeout if one doens't come up in a reasonable amount of time
start on mounted MOUNTPOINT=/ and stopped cloud-init-local
-stop on net-device-up IFACE!=lo
+stop on NEW_NETWORK_
task
console output