New razor installation; won't install

79 views
Skip to first unread message

Bret Wortman

unread,
Apr 12, 2016, 1:20:16 PM4/12/16
to puppet-razor
I'm working on setting up a new razor server and have gotten the microkernel boot to work fine. But now, when it tries to provision the new node, I get this:

:
Filename: bootstrap.ipxe
bootstrap.ipxe : 1626 bytes [script]
Noop installer (Installed by Razor)
forcing local booting with sanboot 0x80
Booting from SAN device 0x80
Boot from SAN device 0x80 failed: Exec format error (http://ipxe.org/2e852001)
Could not boot: Exec format error (http://ipxe.org/2e852001)
Could not boot image: Result too large (http://ipxe.org/46022001)
No more network devices

Operating System not found

Now, I noticed two things here: First, it's using a "Noop" installer when the profile I have calls for centos/7:

[root@puppet ~]# razor policies

+---------+--------+--------+--------+--------+--------+--------+-------+
| name    | repo   | task   | broker | enable | max_co | tags   | nodes |
|         |        |        |        | d      | unt    |        |       |
+---------+--------+--------+--------+--------+--------+--------+-------+
| centos- | centos | centos | puppet | true   |        | (none) | 0     |
| for-sma | 7      | /7     |        |        |        |        |       |
| ll      |        |        |        |        |        |        |       |
+---------+--------+--------+--------+--------+--------+--------+-------+

Query an entry by including its name, e.g. `razor policies centos-for-small`

[root@puppet ~]# razor brokers

+--------+-------------+-----------------------------------------+----------+
| name   | broker_type | configuration                           | policies |
+--------+-------------+-----------------------------------------+----------+
| noop   | noop        | (none)                                  | 0        |
+--------+-------------+-----------------------------------------+----------+
| puppet | puppet      | {"server"=>"puppet.wedgeofli.me", "envi | 1        |
|        |             | ronment"=>"production"}                 |          |
+--------+-------------+-----------------------------------------+----------+

Query an entry by including its name, e.g. `razor brokers noop`

[root@puppet ~]# razor repos centos7

     name: centos7
  iso_url: ---
     task: centos/7

Query additional details via: `razor repos centos7 [task]`

[root@puppet ~]# razor policies centos-for-small

       name: centos-for-small
       repo: centos7
       task: centos/7
     broker: puppet
    enabled: true
  max_count: nil
       tags: (none)
      nodes: 0

Query additional details via: `razor policies centos-for-small [broker, configuration, nodes, repo, tags, task]`

[root@puppet ~]# 

So I'm not sure where the noop is coming from. It had been set in the policy originally, but I updated it to what you see here and still it's trying to use the noop installer.

Secondly, The url http://192.168.1.101:8150/api/nodes/5 is wrong -- it should be /api/collections/nodes/node5. Is this partly to blame, and if so, why is razor using the wrong url to its own api? Did I mis-set this somewhere?

Thanks.

Bret Wortman

unread,
Apr 13, 2016, 9:21:24 AM4/13/16
to puppet-razor
For completeness, I had a repo defined thusly:

[root@puppet repo-store]# razor repos centos7

     name: centos7
  iso_url: ---
     task: centos

Query additional details via: `razor repos centos7 [task]`

[root@puppet repo-store]# 

When I changed to using a local repo like this one, everything worked fine:

[root@puppet repo-store]# razor repos centos7-local

     name: centos7-local
      url: ---
     task: centos

Query additional details via: `razor repos centos7-local [task]`

[root@puppet repo-store]# 

Now I'm getting errors about finding a valid baseurl for repo: base/7/x86_64, but it's still progress -- the box ran the MK, then installed, and only failed after that point. This is pretty cool stuff.


Bret

Scott McClellan

unread,
Apr 15, 2016, 5:25:06 PM4/15/16
to puppet...@googlegroups.com
Hi Bret,

On Wed, Apr 13, 2016 at 8:21 AM, Bret Wortman <br...@thewortmans.org> wrote:
For completeness, I had a repo defined thusly:

[root@puppet repo-store]# razor repos centos7

     name: centos7
  iso_url: ---
     task: centos

Query additional details via: `razor repos centos7 [task]`

[root@puppet repo-store]# 

When I changed to using a local repo like this one, everything worked fine:

[root@puppet repo-store]# razor repos centos7-local

     name: centos7-local
      url: ---
     task: centos

Query additional details via: `razor repos centos7-local [task]`

[root@puppet repo-store]# 

Both methods should work just fine. I was able to complete an install using the `url` argument with that exact URL. Without seeing the node log and/or server log, I can't be sure what happened here.
 

Now I'm getting errors about finding a valid baseurl for repo: base/7/x86_64, but it's still progress -- the box ran the MK, then installed, and only failed after that point. This is pretty cool stuff.

I'm not clear on what you mean here — is this in the broker? Where is this error showing up?
You did this by running the `update-policy-task` command? I have verified that this command does update the policy's task immediately, so I'm not sure what the problem here is.
 
Secondly, The url http://192.168.1.101:8150/api/nodes/5 is wrong -- it should be /api/collections/nodes/node5. Is this partly to blame, and if so, why is razor using the wrong url to its own api? Did I mis-set this somewhere?

That URL isn't referenced anywhere. It will be updated for consistency, but that has no functional effect on your install.


Thanks.

--
You received this message because you are subscribed to the Google Groups "puppet-razor" group.
To unsubscribe from this group and stop receiving emails from it, send an email to puppet-razor...@googlegroups.com.
To post to this group, send email to puppet...@googlegroups.com.
Visit this group at https://groups.google.com/group/puppet-razor.
For more options, visit https://groups.google.com/d/optout.

Bret Wortman

unread,
Apr 15, 2016, 6:58:25 PM4/15/16
to puppet...@googlegroups.com
The second error waddle to nodes "squatting" in our dhcp range. I'll grab logs for the other Monday. Thanks!
You received this message because you are subscribed to a topic in the Google Groups "puppet-razor" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/puppet-razor/pWz5ukgrXA0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to puppet-razor...@googlegroups.com.

Bret Wortman

unread,
Apr 18, 2016, 7:50:37 AM4/18/16
to puppet-razor
"waddle" = "was due to".
Hi Bret,


Thanks.
To unsubscribe from this group and stop receiving emails from it, send an email to puppet-razor+unsubscribe@googlegroups.com.

To post to this group, send email to puppet...@googlegroups.com.
Visit this group at https://groups.google.com/group/puppet-razor.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "puppet-razor" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/puppet-razor/pWz5ukgrXA0/unsubscribe.
To unsubscribe from this group and all its topics, send an email to puppet-razor+unsubscribe@googlegroups.com.

Bret Wortman

unread,
Apr 18, 2016, 9:33:44 AM4/18/16
to puppet-razor
Scott,

I did get this to complete succesfully by defining the repo to be locally hosted. Pointing to the remote CentOS 7 repo simply didn't work for us. Not sure why, but it's not a big deal. I'm turning my attention to figuring out how to get Puppet (or Razor, during installation) to properly assign hostname and IP by looking up the MAC from a database.


Bret
Reply all
Reply to author
Forward
0 new messages