EC2 ap-southeast-2 status failures

[Tim Penhey]

Hi all,

tl;dr: it's not our fault

More debugging... don't we all love debugging live environments.

What I've found is that the query that is used to connect even to the
bootstrap node is failing.  AWS is just not responding correctly.  It
seems that in order to get any form of technical support you need a
support contract with Amazon.

The request is this one (with get params formatted nicely):

https://ec2.ap-southeast-2.amazonaws.com/?
Action=DescribeInstances
AWSAccessKeyId=[redacted]
Filter.1.Name=group-name
Filter.1.Value.1=juju-amazon-ap
Filter.2.Name=instance-id
Filter.2.Value.1=i-697eb454
Filter.3.Name=instance-state-name
Filter.3.Value.1=pending
Filter.3.Value.2=running
Timestamp=2013-04-22T03%3A14%3A46Z
SignatureMethod=HmacSHA256
SignatureVersion=2
Signature=[redacted]
Version=2011-12-15


And the response is something like this:

HTTP/1.1 200 OK
Transfer-Encoding: chunked
Content-Type: text/xml;charset=UTF-8
Date: Mon, 22 Apr 2013 03:14:52 GMT
Server: AmazonEC2
Vary: Accept-Encoding

<?xml version="1.0" encoding="UTF-8"?>
<DescribeInstancesResponse xmlns="http://ec2.amazonaws.com/doc/2011-12-15/">
    <requestId>e083f8e1-d619-4bd2-a8e6-4126153fbd98</requestId>
    <reservationSet/>
</DescribeInstancesResponse>


I can see the image is up and running using the AWS Console, so it is
just likely that the API endpoint is somehow screwed up.

Exactly the same code works fine on us-east-1, with the same type of get
http call, but with lots more xml response data.


How do we move forwards with this?  It certainly looks like an AWS
issue, not a juju issue.

Tim