SSHClient does not properly handle txzookeeper connection errors

Bug #892254 reported by Jim Baker
This bug report is a duplicate of:  Bug #901463: SSH Client code and output cleanups. Edit Remove
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
pyjuju
New
Low
Unassigned

Bug Description

The implementation of SSHClient._internal_connect does not have an effective errback in place, resulting in the errors being handled by the reactor loop, then logged to stderr. An example of this can be seen in logs like this:

Unhandled error in Deferred:
Unhandled Error
Traceback (most recent call last):

from http://wtf.labix.org/413/ec2-wordpress.out

Fixing this bug is subtle/time consuming, and so I'm decoupling it from the related bug 849071. In addition, in the course of looking at this bug, there are a number of related issues in test_sshclient, which verifies the desired sequence of tunnel setup, port watching, and so forth necessary to setup SSHClient. In particular, test_share_connection doesn't currently test anything since it's missing inlineCallbacks (and simply adding that makes the test fail).

Jim Baker (jimbaker)
Changed in juju:
milestone: none → florence
importance: Undecided → Low
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.