search

Alignak-monitoring / alignak-packaging

Packaging files for alignak applications
0 stars 0 forks source link

Unable to restart scheduler and broker on ubuntu 16.04 #28

Open fpeyre opened 7 years ago

fpeyre commented 7 years ago

I install Alignak with the package for ubuntu 16.04 (At the moment when I make this issue : alignak-all_0.2-1.1489300848.c527ae7_all.deb )

I have some issues when i want restart with systemctl the scheduler and the broker

For the scheduler, after systemctl restart alignak-scheduler, I make

systemctl status alignak-scheduler
● alignak-scheduler.service - Alignak Scheduler
   Loaded: loaded (/lib/systemd/system/alignak-scheduler.service; enabled; vendor preset: enabled)
   Active: failed (Result: exit-code) since Thu 2017-03-23 15:14:04 UTC; 48s ago
  Process: 23884 ExecStart=/usr/bin/alignak-scheduler -d -c /etc/alignak/daemons/schedulerd.ini (code=exited, status=1/FAILURE)
 Main PID: 23455 (code=killed, signal=KILL)
   CGroup: /lxc/infra-core-monitoring-test-3064715/system.slice/alignak-scheduler.service
           ├─23456 alignak-scheduler                                                                   
           └─23761 alignak-scheduler-master module: logs                                               

Mar 23 15:14:04 infra-core-monitoring-test-3064715 alignak-scheduler[23884]:     sys.exit(main())
Mar 23 15:14:04 infra-core-monitoring-test-3064715 alignak-scheduler[23884]:   File "/usr/lib/python2.7/dist-packages/alignak/bin/alignak_scheduler.py", line 97, in main
Mar 23 15:14:04 infra-core-monitoring-test-3064715 alignak-scheduler[23884]:     daemon.main()
Mar 23 15:14:04 infra-core-monitoring-test-3064715 alignak-scheduler[23884]:   File "/usr/lib/python2.7/dist-packages/alignak/daemons/schedulerdaemon.py", line 391, in main
Mar 23 15:14:04 infra-core-monitoring-test-3064715 alignak-scheduler[23884]:     self.uri = self.http_daemon.uri
Mar 23 15:14:04 infra-core-monitoring-test-3064715 alignak-scheduler[23884]: AttributeError: 'NoneType' object has no attribute 'uri'
Mar 23 15:14:04 infra-core-monitoring-test-3064715 systemd[1]: alignak-scheduler.service: Control process exited, code=exited status=1
Mar 23 15:14:04 infra-core-monitoring-test-3064715 systemd[1]: Failed to start Alignak Scheduler.
Mar 23 15:14:04 infra-core-monitoring-test-3064715 systemd[1]: alignak-scheduler.service: Unit entered failed state.
Mar 23 15:14:04 infra-core-monitoring-test-3064715 systemd[1]: alignak-scheduler.service: Failed with result 'exit-code'.

I make the same thing for the broker. We have a different error message

systemctl status alignak-broker
● alignak-broker.service - Alignak Broker
   Loaded: loaded (/lib/systemd/system/alignak-broker.service; enabled; vendor preset: enabled)
   Active: failed (Result: timeout) since Thu 2017-03-23 15:21:30 UTC; 16s ago
  Process: 23942 ExecStart=/usr/bin/alignak-broker -d -c /etc/alignak/daemons/brokerd.ini (code=exited, status=0/SUCCESS)
 Main PID: 23414 (code=killed, signal=KILL)
   CGroup: /lxc/infra-core-monitoring-test-3064715/system.slice/alignak-broker.service
           ├─23415 alignak-broker                                                                
           └─23639 alignak-broker-master module: backend_broker                                  

Mar 23 15:20:01 infra-core-monitoring-test-3064715 alignak-broker[23942]:   File "/usr/lib/python2.7/dist-packages/alignak/http/daemon.py", line 122, in __init__
Mar 23 15:20:01 infra-core-monitoring-test-3064715 alignak-broker[23942]:     raise PortNotFree(msg)
Mar 23 15:20:01 infra-core-monitoring-test-3064715 alignak-broker[23942]: PortNotFree: Error: Sorry, the port 0.0.0.0/7772 is not free
Mar 23 15:20:01 infra-core-monitoring-test-3064715 alignak-broker[23942]: [2017-03-23 15:20:00 UTC] INFO: [alignak.daemon] Waiting for initial configuration
Mar 23 15:21:30 infra-core-monitoring-test-3064715 systemd[1]: alignak-broker.service: Start operation timed out. Terminating.
Mar 23 15:21:30 infra-core-monitoring-test-3064715 alignak-broker[23942]: .........................................................................................[2017-03-23 15:21:30 UTC] INFO: [alignak.daemon] process 23942 received a signal: 15
Mar 23 15:21:30 infra-core-monitoring-test-3064715 alignak-broker[23942]: .
Mar 23 15:21:30 infra-core-monitoring-test-3064715 systemd[1]: Failed to start Alignak Broker.
Mar 23 15:21:30 infra-core-monitoring-test-3064715 systemd[1]: alignak-broker.service: Unit entered failed state.
Mar 23 15:21:30 infra-core-monitoring-test-3064715 systemd[1]: alignak-broker.service: Failed with result 'timeout'.

I don't know if it's specific to the package or alignak itself. If I am at the wrong place, i will create the issue on Alignak.

I am available to make more test or give you more information

fpeyre commented 7 years ago

Ok for more information:

My use case is not just a simple restart. I make systemctl start alignak-scheduler once I make a restart, everything seems work I make systemctl start alignak-scheduler I make restart -> We have the error message

I join the log in a file schedulerd.log.txt brokerd.log.txt

mohierf commented 7 years ago

In the scheduler log, the interesting error is:

[2017-03-23 17:34:30 UTC] ERROR: [alignak.daemon] The HTTP daemon port is not free...
[2017-03-23 17:34:30 UTC] ERROR: [alignak.daemon] The HTTP daemon port is not free: Error: Sorry, the port 0.0.0.0/7768 is not free
Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/alignak/daemon.py", line 784, in setup_communication_daemon
    ssl_cert, server_dh, self.daemon_thread_pool_size)
  File "/usr/lib/python2.7/dist-packages/alignak/http/daemon.py", line 122, in __init__
    raise PortNotFree(msg)
PortNotFree: Error: Sorry, the port 0.0.0.0/7768 is not free
[2017-03-23 17:34:30 UTC] CRITICAL: [alignak.daemon] I got an unrecoverable error. I have to exit.
[2017-03-23 17:34:30 UTC] CRITICAL: [alignak.daemon] You can get help at https://github.com/Alignak-monitoring/alignak
[2017-03-23 17:34:30 UTC] CRITICAL: [alignak.daemon] If you think this is a bug, create a new issue including as much details as possible (version, configuration, ..
[2017-03-23 17:34:30 UTC] CRITICAL: [alignak.daemon] -----
[2017-03-23 17:34:30 UTC] CRITICAL: [alignak.daemon] Back trace of the error: Traceback (most recent call last):
  File "/usr/lib/python2.7/dist-packages/alignak/daemons/schedulerdaemon.py", line 391, in main
    self.uri = self.http_daemon.uri
AttributeError: 'NoneType' object has no attribute 'uri'

This problem happens once the scheduler got started and received a configuration. All the previous start/stop of the daemon do not raise any problem.

Note that the broker has the same behaviour 😉

And I think that this may be related with https://github.com/Alignak-monitoring/alignak/issues/673