[Oneiric] Cannot deply services - store.juju.ubuntu.com not found

Bug #872164 reported by Shang Wu
12
This bug affects 2 people
Affects Status Importance Assigned to Milestone
pyjuju
Fix Released
Undecided
Kapil Thangavelu

Bug Description

With the latest updates on Oneiric, when I tried the following command, I got some error message:

ubuntu@ubuntu-Aspire-M5910:~$ juju
error: No environments configured. Please edit: /home/ubuntu/.juju/environments.yaml
ubuntu@ubuntu-Aspire-M5910:~$ cd .juju/
ubuntu@ubuntu-Aspire-M5910:~/.juju$ ls
environments.yaml
ubuntu@ubuntu-Aspire-M5910:~/.juju$ nano environments.yaml
ubuntu@ubuntu-Aspire-M5910:~/.juju$ juju status
juju environment not found: is the environment bootstrapped?
2011-10-11 16:12:05,700 ERROR juju environment not found: is the environment bootstrapped?
ubuntu@ubuntu-Aspire-M5910:~/.juju$ juju bootstrap
2011-10-11 16:15:10,816 INFO Bootstrapping environment 'sample' (type: ec2)...
2011-10-11 16:15:21,503 INFO 'bootstrap' command finished successfully
ubuntu@ubuntu-Aspire-M5910:~/.juju$ juju deploy --repository=examples wordpress
DNS lookup failed: address 'store.juju.ubuntu.com' not found: [Errno -5] No address associated with hostname.
2011-10-11 16:15:50,215 ERROR DNS lookup failed: address 'store.juju.ubuntu.com' not found: [Errno -5] No address associated with hostname.

ubuntu@ubuntu-Aspire-M5910:~/.juju$ dpkg -l juju
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name Version Description
+++-=================================-=================================-==================================================================================
ii juju 0.5+bzr398-0ubuntu1 next generation service orchestration system

Revision history for this message
James Page (james-page) wrote :

Hi Shang

The most recent version of juju in Oneiric introduces changes to support both local and charm store location of services.

You will need to change how you deploy services using the local: prefix with the charm name -

   juju deploy --repository=examples local:wordpress

see https://lists.ubuntu.com/archives/juju/2011-October/000850.html for a few more details about this change.

Marking this bug as invalid as this is an expected break.

Changed in juju:
status: New → Invalid
Revision history for this message
William Reade (fwereade) wrote :

The interface changed a few days ago; you now need to do the following:

ubuntu@ubuntu-Aspire-M5910:~/.juju$ juju deploy --repository=examples local:wordpress

The documentation *should* be up to date now;i f you're aware of any we've missed, please let us know so we can fix it.

Revision history for this message
Shang Wu (shangwu) wrote :

thanks for the info.

Revision history for this message
Shang Wu (shangwu) wrote :

@William,

If you are referring to the documentation here:
https://juju.ubuntu.com/Documentation

I don't think it is up-to-date.

Revision history for this message
Kapil Thangavelu (hazmat) wrote : Re: [Bug 872164] Re: [Oneiric] Cannot deply services - store.juju.ubuntu.com not found

Excerpts from Shang Wu's message of Fri Oct 14 04:24:34 UTC 2011:
> @William,
>
> If you are referring to the documentation here:
> https://juju.ubuntu.com/Documentation
>
> I don't think it is up-to-date.
>

I've filed an RT IS request on this about 30hrs ago, its in the queue to be
fixed.

thanks

Revision history for this message
Kapil Thangavelu (hazmat) wrote :

Excerpts from Shang Wu's message of Fri Oct 14 04:24:34 UTC 2011:
> @William,
>
> If you are referring to the documentation here:
> https://juju.ubuntu.com/Documentation
>
> I don't think it is up-to-date.
>

I've filed an rt about this getting the docs in a cron job updated.

I've tried to communicate the urgency..
https://portal.admin.canonical.com/48456

Otoh, i don't think this code should be attempting to contact the remote store
for an unqualified name if you specify a local disk repository.

The user should not have to express their intention twice.

Revision history for this message
Clint Byrum (clint-fewbar) wrote :

Excerpts from Kapil Thangavelu's message of Sat Oct 15 13:44:30 UTC 2011:
> Excerpts from Shang Wu's message of Fri Oct 14 04:24:34 UTC 2011:
> > @William,
> >
> > If you are referring to the documentation here:
> > https://juju.ubuntu.com/Documentation
> >
> > I don't think it is up-to-date.
> >
>
> I've filed an rt about this getting the docs in a cron job updated.
>
> I've tried to communicate the urgency..
> https://portal.admin.canonical.com/48456
>
>
> Otoh, i don't think this code should be attempting to contact the remote store
> for an unqualified name if you specify a local disk repository.
>
> The user should not have to express their intention twice.
>

Right, no store available should just mean that the namespace resolution
degrades to local charms only.

Is it too strong to say that local charms should always override the
store namespace when the charm name is unqualified? Like if I've written
a replacement mysql, and put it in my repo, when is it ok to replace my
local mysql with the store's?

Ideally I'd be able to set the precedence of all available namespaces.

Revision history for this message
Kapil Thangavelu (hazmat) wrote :

Excerpts from Clint Byrum's message of Sat Oct 15 16:49:54 UTC 2011:
> Excerpts from Kapil Thangavelu's message of Sat Oct 15 13:44:30 UTC 2011:
> > Excerpts from Shang Wu's message of Fri Oct 14 04:24:34 UTC 2011:
> > > @William,
> > >
> > > If you are referring to the documentation here:
> > > https://juju.ubuntu.com/Documentation
> > >
> > > I don't think it is up-to-date.
> > >
> >
> > I've filed an rt about this getting the docs in a cron job updated.
> >
> > I've tried to communicate the urgency..
> > https://portal.admin.canonical.com/48456
> >
> >
> > Otoh, i don't think this code should be attempting to contact the remote store
> > for an unqualified name if you specify a local disk repository.
> >
> > The user should not have to express their intention twice.
> >
>
> Right, no store available should just mean that the namespace resolution
> degrades to local charms only.

disregarding the charm store entirely, if the command line is

juju deploy --repostory foobar

The user has expressed an intent to use a local repository, and the charm should
be looked up from there without the need for a namespace qualifier.

I think that's the issue that should be fixed for this bug.

>
> Is it too strong to say that local charms should always override the
> store namespace when the charm name is unqualified? Like if I've written
> a replacement mysql, and put it in my repo, when is it ok to replace my
> local mysql with the store's?

>
> Ideally I'd be able to set the precedence of all available namespaces.
>

multi repo usage here is a little interesting, and at conflicts with gustavo's
notion (expressed at capetown sprint) of a charm store doing full dep
graph resolution. multi-repo usage needs client side dep resolution with
repo precendents. there's an implementation of client side auto dep
resolution with precedence i wrote in trunk history.

the auto resolve deps precendence is an ordered list of repos, with the dep
installation from the repo that last successfully looked up as the default start
for the next dep, and then falling back to the start of the ordered repo list.

Changed in juju:
assignee: nobody → Kapil Thangavelu (hazmat)
milestone: none → florence
status: Invalid → In Progress
Revision history for this message
Kapil Thangavelu (hazmat) wrote :

Also to clarify multi-repo in this case would be both a local repo and a charm store repo, multiple namespaces in the charm store would definitely be supported server side.

Revision history for this message
Kapil Thangavelu (hazmat) wrote :

it seems the best option that will pass muster is to just log the repo being utilized for the lookup since local repository usage needs to be spelled twice on the cli.

Revision history for this message
Kapil Thangavelu (hazmat) wrote :

the store is almost live.. marking won't fix

Changed in juju:
status: In Progress → Fix Released
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.