[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: Origin 4 problem to add nodes to district

2015-01-14 17:42 GMT+00:00 Miciah Dashiel Butler Masters <mmasters redhat com>:
On Wed, Jan 14, 2015 at 06:25:21PM +0100, Massimiliano Dessì wrote:
2015-01-14 17:55 GMT+01:00 Miciah Dashiel Butler Masters <
mmasters redhat com>:

On Wed, Jan 14, 2015 at 01:55:29PM +0000, Massimiliano Dessě wrote:

HI all,
the oo-mco ping show  the two nodes,
but when I try to add with
oo-admin-ctl-district -c add-node -n small_district -i <node>
I receive the message
"ERROR OUTPUT: Cannot connect to node."
Is /opt/rh/ruby193/root/etc/mcollective/facts.yaml being updated? The
/etc/cron.minutely/openshift-facts cronjob should update facts.yaml
once every minute.  If facts.yaml isn't being updated, try running
`oo-exec-ruby /opt/rh/ruby193/root/usr/libexec/mcollective/update_yaml.rb`
manually to see whether that fails or whether cron is misconfigured.

yes is updated every minute

And the contents look OK? Do facts such as kernel, node_profile, and
public_hostname look correct?

 The oo-mco facts node-profile -v

That should be 'node_profile', with an underscore.

[root broker mcollective]# oo-mco facts node_profile -v
Discovering hosts using the mc method for 2 second(s) .... 2
Report for fact: node_profile

---- rpc stats ----
          Nodes: 2 / 0
    Pass / Fail: 0 / 0
     Start Time: 2015-01-14 09:15:00 -0800
 Discovery Time: 2050.83ms
     Agent Time: 12005.06ms
     Total Time: 14055.89ms

No response from:

  node01.<mydomain>  node02.<mydomain>

 show a message no response from the two nodes.

Any hints to fix ?

What does `oo-mco facts kernel` show? Should be 'Linux' (same as the
output of `uname -s`).

[root broker ~]# uname -s
[root broker mcollective]# oo-mco facts kernel
Report for fact: kernel

Finished processing 0 / 2 hosts in 12004.92 ms

No response from:

  node01.<mydomain>  node02.<mydomain>

How about `oo-mco inventory <node>`?

[root broker mcollective]# oo-mco inventory node01.<mydomain>
Did not receive any results from node node01.<mydomain>

Is oo-mco ping consistently showing responses from both nodes? If there
are intermittent failures, maybe the hosts' clocks are out of synch, or
there could be other problems with the ActiveMQ connections. 

It's all ok on the nodes
It looks
like MCollective is authenticating with ActiveMQ fine though and has the
correct collective configured in
/opt/rh/ruby193/root/etc/mcollective/server.cfg and
/opt/rh/ruby193/root/etc/mcollective/client.cfg seeing that oo-mco ping
is working.

Do you have direct_addressing = 1 in server.cfg and direct_addressing
= 0 in client.cfg? If not, try with those settings (you'll need to
restart ruby193-mcollective on the nodes to test this, and you'll need
to restart openshift-broker on the broker host if testing shows that
setting to be the issue).

Bingo !
I have the direct_addressing = 0 on the server.cfg
and nothing in the client.cfg,
but with direct_addressing=0 in the server and
with direct_addressing=1 in the client.cfg I'm able to add the node with

oo-admin-ctl-district -c add-node -n small_district -i <node>

why in the comprehensive guide is not present a check about this flags if the default values are wrong ?

Could you tailf /var/log/openshift/node/ruby193-mcollective.log on the
nodes as you run the oo-mco inventory or oo-mco facts commands? That
will let us determine whether the node is ignoring the request, on for
what reason.  I'm attaching some example log output in case that might
help you understand what I am asking you to look for.

Miciah Dashiel Butler Masters <mmasters redhat com>

Thanks Dashiel !

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]