BotoServerError: 505 HTTP Version Not Supported

1,000 views
Skip to first unread message

Keith Fitzgerald

unread,
Feb 17, 2011, 4:00:08 PM2/17/11
to boto...@googlegroups.com
I'm suddenly getting the error below while creating an autoscale group. I'm still running 2.0b4 of boto so I'm not sure why this is suddenly failing. 

Has anyone encountered this before? Is this an amazon issue? 

Traceback (most recent call last):
  File "cycle-instance.py", line 138, in <module>
    as_connection.create_launch_configuration(launch_config)
  File "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/ec2/autoscale/__init__.py", line 136, in create_launch_configuration
    Request)
  File "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/connection.py", line 611, in get_object
    raise self.ResponseError(response.status, response.reason, body)
boto.exception.BotoServerError: BotoServerError: 505 HTTP Version Not Supported

Keith Fitzgerald

unread,
Feb 17, 2011, 4:01:10 PM2/17/11
to boto...@googlegroups.com
well that was fast. the issue resolved it self. must have been an amazon issue after all.

Mitchell Garnaat

unread,
Feb 17, 2011, 4:03:21 PM2/17/11
to boto...@googlegroups.com
Hmm.  Never saw that one before.  I think it must have been a transient config error on the AWS frontend servers.

Mitch

Keith Fitzgerald

unread,
Feb 22, 2011, 4:39:30 PM2/22/11
to boto...@googlegroups.com
Awesome. Thanks Mitch.

On Tuesday, February 22, 2011 at 4:11 PM, Mitchell Garnaat wrote:

The best way is to follow this incantation:

import boto

boto.set_stream_logger('foo')
ec2 = boto.connect_ec2(debug=2)

Then you should get copious debug output to your command window.

Mitch

On Tue, Feb 22, 2011 at 3:10 PM, Keith Fitzgerald <kfitz...@gmail.com> wrote:
not yet, trying to pull out the actual AWS error. is there a way to put boto in debug mode to dump AWS errors? 

On Tuesday, February 22, 2011 at 3:02 PM, Mitchell Garnaat wrote:

Have you posted anything to the AWS forums or searched there for previous posts on the issue?

Mitch

On Tue, Feb 22, 2011 at 2:43 PM, ghostrocket <kfitz...@gmail.com> wrote:
grrr. this problem is happening again. it happens when creating an
autoscale group ... has anyone else experienced this issue?


On Feb 17, 4:03 pm, Mitchell Garnaat <mi...@garnaat.com> wrote:
> Hmm.  Never saw that one before.  I think it must have been a transient
> config error on the AWS frontend servers.
>
> Mitch
>
> On Thu, Feb 17, 2011 at 4:01 PM, Keith Fitzgerald <kfitzger...@gmail.com>wrote:
>
>
>
> >  well that was fast. the issue resolved it self. must have been an amazon
> > issue after all.
>
> > On Thursday, February 17, 2011 at 4:00 PM, Keith Fitzgerald wrote:
>
> >   I'm suddenly getting the error below while creating an autoscale group.
> > I'm still running 2.0b4 of boto so I'm not sure why this is suddenly
> > failing.
>
> > Has anyone encountered this before? Is this an amazon issue?
>
> > Traceback (most recent call last):
> >   File "cycle-instance.py", line 138, in <module>
> >     as_connection.create_launch_configuration(launch_config)
> >   File
> > "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/e c2/autoscale/__init__.py",

> > line 136, in create_launch_configuration
> >     Request)
> >   File
> > "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/c onnection.py",

Keith Fitzgerald

unread,
Feb 22, 2011, 3:10:55 PM2/22/11
to boto...@googlegroups.com
not yet, trying to pull out the actual AWS error. is there a way to put boto in debug mode to dump AWS errors? 

On Tuesday, February 22, 2011 at 3:02 PM, Mitchell Garnaat wrote:

Have you posted anything to the AWS forums or searched there for previous posts on the issue?

Mitch

On Tue, Feb 22, 2011 at 2:43 PM, ghostrocket <kfitz...@gmail.com> wrote:
grrr. this problem is happening again. it happens when creating an
autoscale group ... has anyone else experienced this issue?

On Feb 17, 4:03 pm, Mitchell Garnaat <mi...@garnaat.com> wrote:
> Hmm.  Never saw that one before.  I think it must have been a transient
> config error on the AWS frontend servers.
>
> Mitch
>
> On Thu, Feb 17, 2011 at 4:01 PM, Keith Fitzgerald <kfitzger...@gmail.com>wrote:
>
>
>
> >  well that was fast. the issue resolved it self. must have been an amazon
> > issue after all.
>
> > On Thursday, February 17, 2011 at 4:00 PM, Keith Fitzgerald wrote:
>
> >   I'm suddenly getting the error below while creating an autoscale group.
> > I'm still running 2.0b4 of boto so I'm not sure why this is suddenly
> > failing.
>
> > Has anyone encountered this before? Is this an amazon issue?
>
> > Traceback (most recent call last):
> >   File "cycle-instance.py", line 138, in <module>
> >     as_connection.create_launch_configuration(launch_config)
> >   File
> > "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/e c2/autoscale/__init__.py",

> > line 136, in create_launch_configuration
> >     Request)
> >   File
> > "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/c onnection.py",

ghostrocket

unread,
Feb 22, 2011, 2:43:05 PM2/22/11
to boto-dev
grrr. this problem is happening again. it happens when creating an
autoscale group ... has anyone else experienced this issue?

On Feb 17, 4:03 pm, Mitchell Garnaat <mi...@garnaat.com> wrote:
> Hmm.  Never saw that one before.  I think it must have been a transient
> config error on the AWS frontend servers.
>
> Mitch
>
> On Thu, Feb 17, 2011 at 4:01 PM, Keith Fitzgerald <kfitzger...@gmail.com>wrote:
>
>
>
> >  well that was fast. the issue resolved it self. must have been an amazon
> > issue after all.
>
> > On Thursday, February 17, 2011 at 4:00 PM, Keith Fitzgerald wrote:
>
> >   I'm suddenly getting the error below while creating an autoscale group.
> > I'm still running 2.0b4 of boto so I'm not sure why this is suddenly
> > failing.
>
> > Has anyone encountered this before? Is this an amazon issue?
>
> > Traceback (most recent call last):
> >   File "cycle-instance.py", line 138, in <module>
> >     as_connection.create_launch_configuration(launch_config)
> >   File
> > "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/e c2/autoscale/__init__.py",
> > line 136, in create_launch_configuration
> >     Request)
> >   File
> > "/Users/ghostrocket/Projects/pinup/.venv/lib/python2.6/site-packages/boto/c onnection.py",

Mitchell Garnaat

unread,
Feb 22, 2011, 4:11:17 PM2/22/11
to boto...@googlegroups.com
The best way is to follow this incantation:

import boto

boto.set_stream_logger('foo')
ec2 = boto.connect_ec2(debug=2)

Then you should get copious debug output to your command window.

Mitch

Mitchell Garnaat

unread,
Feb 22, 2011, 3:02:08 PM2/22/11
to boto...@googlegroups.com
Have you posted anything to the AWS forums or searched there for previous posts on the issue?

Mitch

On Tue, Feb 22, 2011 at 2:43 PM, ghostrocket <kfitz...@gmail.com> wrote:

Keith Fitzgerald

unread,
Feb 23, 2011, 4:56:59 PM2/23/11
to boto...@googlegroups.com
Good news. I found the cause of this error and committed a fix. https://github.com/boto/boto/commit/0e5fb3ec0aa886bc8155e94d5b91ab361da09104

The user-data was not being base64 encoded and was being sent to Amazon via GET instead of POST.

Thanks for the tip on the debug flag Mitch!
Reply all
Reply to author
Forward
0 new messages