Comment 8 for bug 1656150

Revision history for this message
Luca Cervigni (cervigni) wrote : Re: [Bug 1656150] Re: Mcollective fails to start in bootstrap of baremetal servers

This is the output of "bash -x /etc/rc.local"

root@bootstrap:~# bash -x /etc/rc.local
+ fix-configs-on-startup
Failed to retrieve unit: Connection timed out
Failed to execute operation: Connection timed out
+ true
+ rm -f /etc/nailgun-agent/nodiscover

Waited more than 5 minutes but nailgun does not restart mcollective.

root@bootstrap:~# service mcollective status
● mcollective.service - The Marionette Collective
    Loaded: loaded (/lib/systemd/system/mcollective.service; enabled;
vendor preset: enabled)
    Active: inactive (dead)

Jan 18 08:05:35 bootstrap systemd[1]: mcollective.service: Trying to
enqueue job mcollective.service/restart/replace
Jan 20 00:35:43 bootstrap systemd[1]: mcollective.service: Trying to
enqueue job mcollective.service/start/replace
Jan 20 00:38:20 bootstrap systemd[1]: mcollective.service: Failed to
send unit change signal for mcollective.service: Transport endpoint is
not connected
Jan 20 00:39:04 bootstrap systemd[1]: mcollective.service: Failed to
send unit change signal for mcollective.service: Transport endpoint is
not connected

Running /etc/rc.local though enabled me to start manually mcollective, a
great deal for my deployment for the moment.

root@bootstrap:~# service mcollective start
root@bootstrap:~#

[root@fuel ~]# mco ping
7 time=61.62 ms
10 time=62.61 ms
11 time=63.48 ms
12 time=64.41 ms
13 time=65.31 ms
9 time=66.21 ms
8 time=67.11 ms
6 time=67.91 ms
2 time=68.84 ms
master time=69.70 ms
1 time=74.22 ms

On 18/01/17 18:45, Georgy Kibardin wrote:
> Does running /etc/rc.local manually afterwards fixes the problem?
> It must fix mcollective config, unmask mcollective.service and then in ~30 sec. nailgun agent must successfully start mcollective. Also /etc/send2syslog.conf must appear.
> To understand where it fails I think it would be usefule to run it as bash -x /etc/rc.local.
>