juju ssh fails on local provider: pty-allocation-request-failed-on-channel-0
Bug #903318 reported by
Mark Mims
This bug affects 3 people
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
pyjuju |
Fix Released
|
High
|
Unassigned | ||
lxc (Ubuntu) |
Fix Released
|
High
|
Unassigned |
Bug Description
juju ssh fails using a local provider.
Using juju with a local provider,
hawk:~ $ cat .juju/environme
default: gmail
environments:
gmail:
<cut out>
demo:
<cut out>
local:
type: local
data-dir: /home/mmm/
hawk:~ $ juju bootstrap -elocal
hawk:~ $ juju deploy -elocal --repository . local:mysql
but when the service is "up", I can not ssh to the machine
hawk:~ $ juju ssh -elocal mysql/0
fails with pty-allocation-
Changed in lxc (Ubuntu): | |
importance: | Undecided → High |
Changed in lxc (Ubuntu): | |
status: | Incomplete → Fix Released |
Changed in juju: | |
status: | Confirmed → Fix Released |
To post a comment you must log in.
Workaround: clear the lxc cache
First, destroy any local juju environment you might have running, then 'rm -Rf /var/cache/lxc/'
I additionally purged packages but suspect the cache clear is all that's necessary. I'll test this out next time around.
Remember that the first lxc deploy after a cache wipe will take a long time while it's rebuilt.
This usually gets me another two weeks or so of heavy local provider usage before the problem creeps up again.