[MAAS] Deploy a charm into machine 0 = log loop
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
juju-core |
New
|
Undecided
|
Unassigned |
Bug Description
Dear JuJu Team,
The last two days I deployed some charms into a small cluster of 4 servers.
In order to avoid using a 24 threads for only managing instances with JuJu I decided to deploy some services (my own GlusterFS charm) into machine 0.
I checked the debug-log and what I've seen is that the log is appended with old messages again and again.
I opened a terminal into machine 0 to check the logs and what I've seen confirmed my first idea : The file -> /var/log/
I decided to remove this file and create a nice trip, destination unknown, for the log messages : ln -s /dev/null all-machines.log.
I am using the latest juju-core from a raring desktop to deploy charms into precise servers with MAAS. I also upgraded MAAS with proposed repository in order to avoid bug #1204507.
Hope this helps to make this project even better for the good of the community !
Kind Regards,
David
description: | updated |
this sounds like a duplicate of bug #1211147. You mention "using the latest" can you run "juju version" to check what version you are actually running?
You can get the newest stable series from "ppa:juju/stable" though that doesn't have this fix (I believe). You would want "ppa:juju/devel" to test if this bug is fixed for you (it *should* be fixed in 1.13.2 but we've had some trouble reproducing and isolating the bug).