icinga2 icon indicating copy to clipboard operation
icinga2 copied to clipboard
verified publisher icon Icinga

Timeperiod exclusions doesn't work as expected

Open darkweaver87 opened this issue 6 years ago • 23 comments

Describe the bug

In our config, we use two timeperiods, one one excluding the other one and it doesn't seem to work when a time period contains multiple sub-timeperiod.

To Reproduce

Here is a configuration sample:

object TimePeriod "24x7" {
 ranges = {
     tuesday = "00:00-23:59"
 }
}

object TimePeriod "onDuty" {
   // object attributes
   excludes = []
   ranges = {
     tuesday = "09:00-09:30"
   }
}

object TimePeriod "offDuty" {
 excludes = ["onDuty"]
 ranges = {
     tuesday = "00:00-23:59"
 }
}

object User "me" {
   email = "[email protected]"
   enable_notifications = true
   groups = []
   period = "24x7"
}

object Host "test" {
   address = "127.0.0.1"
   zone = "fr"

   check_command = "check_host_alive"
   check_interval = 20s
   check_period = "offDuty"
   enable_active_checks = 1
   enable_event_handler = 0
   enable_flapping = 0
   enable_notifications = 0
   enable_passive_checks = 0
   max_check_attempts = 3
   retry_interval = 5s
}

apply Service "test-me" {
 host_name = "test"

 display_name = "test"

 check_command = "dummy"

 check_interval = 20s
 retry_interval = 5s
 check_period = "offDuty"

 enable_notifications = 1

 vars.dummy_state = 2
 vars.dummy_text = "TEST KO"
 vars.is_test = "True"
 vars.notification_interval = 60
 vars.notification_options_states = [Critical, Unknown, Warning]
 vars.notification_options_types = [Problem, Recovery]
 vars.notification_period = "offDuty"

 assign where host.name == "test"
}

apply Notification "service-bysms" to Service {
 // object attributes
 command = "notify-service-by-sms"
 period = "24x7"

 users = ["me"]

 states = service.vars.notification_options_states
 types = service.vars.notification_options_types
 if (service.vars["notification_interval"]) {
         interval = service.vars["notification_interval"]
 }  else  {
     interval = 1800
 }

 if (service.vars["notification_period"]) {
         period = service.vars["notification_period"]
 }

 assign where host.name == "test"
}

Expected behavior

To me, in this example, the offDuty period should be 00:00-09:00,09:30-00:00

Your Environment

Include as many relevant details about the environment you experienced the problem in

  • Version used (icinga2 --version):
icinga2 - The Icinga 2 network monitoring daemon (version: r2.10.5-1)

Copyright (c) 2012-2019 Icinga GmbH (https://icinga.com/)
License GPLv2+: GNU GPL version 2 or later <http://gnu.org/licenses/gpl2.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.

System information:
  Platform: Debian GNU/Linux
  Platform version: 9 (stretch)
  Kernel: Linux
  Kernel version: 4.14.119
  Architecture: x86_64

Build information:
  Compiler: GNU 6.3.0
  Build host: cb654124b660

Application information:

General paths:
  Config directory: /etc/icinga2
  Data directory: /var/lib/icinga2
  Log directory: /var/log/icinga2
  Cache directory: /var/cache/icinga2
  Spool directory: /var/spool/icinga2
  Run directory: /run/icinga2

Old paths (deprecated):
  Installation root: /usr
  Sysconf directory: /etc
  Run directory (base): /run
  Local state directory: /var

Internal paths:
  Package data directory: /usr/share/icinga2
  State path: /var/lib/icinga2/icinga2.state
  Modified attributes path: /var/lib/icinga2/modified-attributes.conf
  Objects path: /var/cache/icinga2/icinga2.debug
  Vars path: /var/cache/icinga2/icinga2.vars
  PID path: /run/icinga2/icinga2.pid
  • Operating System and version:
# lsb_release -a
No LSB modules are available.
Distributor ID:	Debian
Description:	Debian GNU/Linux 9.9 (stretch)
Release:	9.9
Codename:	stretch
  • Enabled features (icinga2 feature list):
Disabled features: command compatlog debuglog elasticsearch gelf graphite influxdb opentsdb perfdata statusdata syslog
Enabled features: api checker ido-pgsql livestatus mainlog notification
  • Config validation (icinga2 daemon -C):
...
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 5293 Services.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 2 LivestatusListeners.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 1 IcingaApplication.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 2070 Hosts.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 1 EventCommand.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 1 FileLogger.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 6 NotificationCommands.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 22087 Notifications.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 1 NotificationComponent.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 154 HostGroups.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 1 ApiListener.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 1 CheckerComponent.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 4 Zones.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 6 Endpoints.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 4 ApiUsers.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 15 Users.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 247 CheckCommands.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 1 IdoPgsqlConnection.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 94 UserGroups.
[2019-08-06 12:51:25 +0200] information/ConfigItem: Instantiated 14 TimePeriods.
[2019-08-06 12:51:25 +0200] information/ScriptGlobal: Dumping variables to file '/var/cache/icinga2/icinga2.vars'
[2019-08-06 12:51:25 +0200] information/cli: Finished validating the configuration file(s).

Thanks !

darkweaver87 avatar Aug 06 '19 11:08 darkweaver87

Do you have the debug logs from Thursday 09:00 til 10:00 to help mitigate the problem? Obviously the TimePeriod object is used as check_period, not notification period, so trace the scheduled checks and their results.

dnsmichi avatar Aug 07 '19 13:08 dnsmichi

Thanks for your reply. Hum unfortunately no. I will try to get some logs today but I will have to spaw a dedicated instance. The timeperiod is used as check and notification period, look at period = service.vars["notification_period"] in the notification apply rule. By the way, do you have a smarter way of migrating notification_period, notification_interval and notification_options from a nagios object (I mean something that can be automated) ?

darkweaver87 avatar Aug 08 '19 06:08 darkweaver87

Here is a configuration which reproduce the problem on a fresh icinga install on debian stretch and v2.10.5-1 (same environment as my production):

# cat /etc/icinga2/conf.d/test.conf 
/*object TimePeriod "24x7" {
 ranges = {
     monday = "00:00-23:59"
     tuesday = "00:00-23:59"
     wednesday = "00:00-23:59"
     thursday = "00:00-23:59"
     friday = "00:00-23:59"
     saturday = "00:00-23:59"
     sunday = "00:00-23:59"
 }
}*/

object TimePeriod "onDuty" {
   // object attributes
   excludes = []
   ranges = {
     monday = "09:00-10:00"
     tuesday = "09:00-10:00"
     wednesday = "09:00-10:00"
     thursday = "09:00-10:00"
     friday = "09:00-10:00"
     saturday = "09:00-10:00"
     sunday = "09:00-10:00"
   }
}

object TimePeriod "offDuty" {
 excludes = ["onDuty"]
 ranges = {
     monday = "00:00-23:59"
     tuesday = "00:00-23:59"
     wednesday = "00:00-23:59"
     thursday = "00:00-23:59"
     friday = "00:00-23:59"
     saturday = "00:00-23:59"
     sunday = "00:00-23:59"
 }
}

object User "me" {
   email = "[email protected]"
   enable_notifications = true
   groups = []
   period = "24x7"
}

object Host "test" {
   address = "127.0.0.1"

   check_command = "hostalive"
   check_interval = 20s
   check_period = "offDuty"
   enable_active_checks = 1
   enable_event_handler = 0
   enable_flapping = 0
   enable_notifications = 0
   enable_passive_checks = 0
   max_check_attempts = 3
   retry_interval = 5s
}

apply Service "test-me" {
 host_name = "test"

 display_name = "test"

 check_command = "dummy"

 check_interval = 20s
 retry_interval = 5s
 check_period = "offDuty"

 enable_notifications = 1

 vars.dummy_state = 2
 vars.dummy_text = "TEST KO"
 vars.is_test = "True"
 vars.notification_interval = 60
 vars.notification_options_states = [Critical, Unknown, Warning]
 vars.notification_options_types = [Problem, Recovery]
 vars.notification_period = "offDuty"

 assign where host.name == "test"
}

apply Notification "service-bysms" to Service {
 // object attributes
 command = "mail-service-notification"
 period = "24x7"

 users = ["me"]

 states = service.vars.notification_options_states
 types = service.vars.notification_options_types
 if (service.vars["notification_interval"]) {
         interval = service.vars["notification_interval"]
 }  else  {
     interval = 1800
 }

 if (service.vars["notification_period"]) {
         period = service.vars["notification_period"]
 }

 assign where host.name == "test"
}

Logs:

[2019-08-08 09:41:04 +0200] debug/CheckerComponent: Scheduling info for checkable 'test!test-me' (2019-08-08 09:41:04 +0200): Object 'test!test-me', Next Check: 2019-08-08 09:41:04 +0200(1.56525e+09).
[2019-08-08 09:41:04 +0200] debug/CheckerComponent: Executing check for 'test!test-me'
[2019-08-08 09:41:04 +0200] debug/Checkable: Update checkable 'test!test-me' with check interval '20' from last check time at 2019-08-08 09:40:43 +0200 (1.56525e+09) to next check time at 2019-08-08 09:41:22 +0200(1.56525e+09).
[2019-08-08 09:41:04 +0200] debug/Checkable: Update checkable 'test!test-me' with check interval '20' from last check time at 2019-08-08 09:41:04 +0200 (1.56525e+09) to next check time at 2019-08-08 09:41:22 +0200(1.56525e+09).
[2019-08-08 09:41:04 +0200] debug/DbEvents: add checkable check history for 'test!test-me'
[2019-08-08 09:41:04 +0200] debug/CheckerComponent: Check finished for object 'test!test-me'
[2019-08-08 09:41:05 +0200] notice/NotificationComponent: Attempting to send reminder notification 'test!test-me!service-bysms'
[2019-08-08 09:41:05 +0200] notice/Notification: Attempting to send reminder notifications for notification object 'test!test-me!service-bysms'.
[2019-08-08 09:41:05 +0200] information/Notification: Sending reminder 'Problem' notification 'test!test-me!service-bysms' for user 'me'
[2019-08-08 09:41:05 +0200] debug/DbEvents: add notification history for 'test!test-me'
[2019-08-08 09:41:05 +0200] debug/DbEvents: add contact notification history for service 'test!test-me' and user 'me'.
[2019-08-08 09:41:05 +0200] notice/Process: Running command '/etc/icinga2/scripts/mail-service-notification.sh' '-4' '127.0.0.1' '-6' '' '-b' '' '-c' '' '-d' '2019-08-08 09:41:05 +0200' '-e' 'test-me' '-l' 'test' '-n' 'test' '-o' 'TEST KO' '-r' '[email protected]' '-s' 'CRITICAL' '-t' 'PROBLEM' '-u' 'test': PID 8820
[2019-08-08 09:41:05 +0200] debug/DbEvents: add log entry history for 'test!test-me'
[2019-08-08 09:41:05 +0200] information/Notification: Completed sending 'Problem' notification 'test!test-me!service-bysms' for checkable 'test!test-me' and user 'me'.
[2019-08-08 09:41:05 +0200] notice/Process: PID 8820 ('/etc/icinga2/scripts/mail-service-notification.sh' '-4' '127.0.0.1' '-6' '' '-b' '' '-c' '' '-d' '2019-08-08 09:41:05 +0200' '-e' 'test-me' '-l' 'test' '-n' 'test' '-o' 'TEST KO' '-r' '[email protected]' '-s' 'CRITICAL' '-t' 'PROBLEM' '-u' 'test') terminated with exit code 0

The notification period through service vars is working correctly:

# icinga2 object list --type Notification --name 'test!test-me!service-bysms'
Object 'test!test-me!service-bysms' of type 'Notification':
  % declared in '/etc/icinga2/conf.d/test.conf', lines 86:1-86:45
  * __name = "test!test-me!service-bysms"
  * command = "mail-service-notification"
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 88:2-88:38
  * command_endpoint = ""
  * host_name = "test"
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 86:1-86:45
  * interval = 60
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 96:10-96:57
  * name = "service-bysms"
  * package = "_etc"
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 86:1-86:45
  * period = "offDuty"
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 89:2-89:16
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 102:10-102:53
  * service_name = "test-me"
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 86:1-86:45
  * source_location
    * first_column = 1
    * first_line = 86
    * last_column = 45
    * last_line = 86
    * path = "/etc/icinga2/conf.d/test.conf"
  * states = [ "Critical", "Unknown", "Warning" ]
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 93:2-93:50
  * templates = [ "service-bysms" ]
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 86:1-86:45
  * times = null
  * type = "Notification"
  * types = [ "Problem", "Recovery" ]
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 94:2-94:48
  * user_groups = null
  * users = [ "me" ]
    % = modified in '/etc/icinga2/conf.d/test.conf', lines 91:2-91:15
  * vars = null
  * zone = ""

To me, in this example, the notification should not have occurred between 9h and 10h. I also tried to play with the prefer_includes attribute but it doesn't change anything :-/

darkweaver87 avatar Aug 08 '19 07:08 darkweaver87

I'm a friend of keeping the details inside the notification objects and apply rules, I don't like the old way of stashing everything into hosts and services for notifications. That's also why the notification objects exist, I am one of the architects of that feature. Anyhow, your approach looks sufficient to me, I didn't understand its intention up until now though.

Logs will help to mitigate further. You can also create a dummy TimePeriod configuration with the above excludes and then query the REST API at /v1/objects/timeperiods for the is_inside attribute at that very moment, e.g. with a loop checking that every second.

dnsmichi avatar Aug 08 '19 07:08 dnsmichi

Yes I understand :-) For now I need my old prod to work along with Icinga in order to validate everything is working as expected. That's why I did this "bad thing" to make my migration script working. Once Icinga will have replaced my old monitoring infrastructure I'll rewrite my config the Icinga-way :-) OK for the is_inside test, will do.

darkweaver87 avatar Aug 08 '19 10:08 darkweaver87

OK so very stange :-) I manage to reproduce it serveral times. Here is the scenario:

  1. add the above configuration (change onDuty timeperiod times for the test purpose)
  2. start Icinga -> it won't work
  3. enable API and restart Icinga
  4. it works! I have the following logs:
[2019-08-08 14:03:32 +0200] notice/CheckerComponent: Skipping check for object 'test!test-me': not in check period 'offDuty'
[2019-08-08 14:03:32 +0200] debug/CheckerComponent: Checks for checkable 'test!test-me' are disabled. Rescheduling check.
[2019-08-08 14:03:32 +0200] debug/Checkable: Update checkable 'test!test-me' with check interval '20' from last check time at 2019-08-08 14:01:39 +0200 (1.56527e+09) to next check time at 2019-08-08 14:03:51 +0200(1.56527e+09).

Timeperiod logs:

  • with API disabled:
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:00:23 2019' <-> 'Fri Aug  9 14:00:23 2019' from TimePeriod '24x7'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Adding segment 'Thu Aug  8 00:00:00 2019' <-> 'Fri Aug  9 00:00:00 2019' to TimePeriod '24x7'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Adding segment 'Fri Aug  9 00:00:00 2019' <-> 'Sat Aug 10 00:00:00 2019' to TimePeriod '24x7'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:00:23 2019' <-> 'Fri Aug  9 14:00:23 2019' from TimePeriod 'never'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:00:23 2019' <-> 'Fri Aug  9 14:00:23 2019' from TimePeriod 'offDuty'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Adding segment 'Thu Aug  8 00:00:00 2019' <-> 'Thu Aug  8 23:59:00 2019' to TimePeriod 'offDuty'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Adding segment 'Fri Aug  9 00:00:00 2019' <-> 'Fri Aug  9 23:59:00 2019' to TimePeriod 'offDuty'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Merge TimePeriod 'offDuty' with 'onDuty' Method: exclude
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 12:35:00 2019' <-> 'Thu Aug  8 13:00:00 2019' from TimePeriod 'offDuty'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Removing segment 'Fri Aug  9 09:00:00 2019' <-> 'Fri Aug  9 10:00:00 2019' from TimePeriod 'offDuty'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:00:23 2019' <-> 'Fri Aug  9 14:00:23 2019' from TimePeriod 'onDuty'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Adding segment 'Thu Aug  8 12:35:00 2019' <-> 'Thu Aug  8 14:05:00 2019' to TimePeriod 'onDuty'
[2019-08-08 14:00:23 +0200] debug/TimePeriod: Adding segment 'Fri Aug  9 09:00:00 2019' <-> 'Fri Aug  9 10:00:00 2019' to TimePeriod 'onDuty'
  • with API enabled:
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:03:14 2019' <-> 'Fri Aug  9 14:03:14 2019' from TimePeriod 'onDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Adding segment 'Thu Aug  8 12:35:00 2019' <-> 'Thu Aug  8 14:05:00 2019' to TimePeriod 'onDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Adding segment 'Fri Aug  9 09:00:00 2019' <-> 'Fri Aug  9 10:00:00 2019' to TimePeriod 'onDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:03:14 2019' <-> 'Fri Aug  9 14:03:14 2019' from TimePeriod 'never'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:03:14 2019' <-> 'Fri Aug  9 14:03:14 2019' from TimePeriod 'offDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Adding segment 'Thu Aug  8 00:00:00 2019' <-> 'Thu Aug  8 23:59:00 2019' to TimePeriod 'offDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Adding segment 'Fri Aug  9 00:00:00 2019' <-> 'Fri Aug  9 23:59:00 2019' to TimePeriod 'offDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Merge TimePeriod 'offDuty' with 'onDuty' Method: exclude
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 12:35:00 2019' <-> 'Thu Aug  8 14:05:00 2019' from TimePeriod 'offDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Removing segment 'Fri Aug  9 09:00:00 2019' <-> 'Fri Aug  9 10:00:00 2019' from TimePeriod 'offDuty'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Removing segment 'Thu Aug  8 14:03:14 2019' <-> 'Fri Aug  9 14:03:14 2019' from TimePeriod '24x7'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Adding segment 'Thu Aug  8 00:00:00 2019' <-> 'Fri Aug  9 00:00:00 2019' to TimePeriod '24x7'
[2019-08-08 14:03:14 +0200] debug/TimePeriod: Adding segment 'Fri Aug  9 00:00:00 2019' <-> 'Sat Aug 10 00:00:00 2019' to TimePeriod '24x7'
[2019-08-08 14:08:14 +0200] debug/TimePeriod: Removing segment 'Fri Aug  9 14:03:14 2019' <-> 'Fri Aug  9 14:08:14 2019' from TimePeriod 'never'
[2019-08-08 14:08:14 +0200] debug/TimePeriod: Removing segment 'Fri Aug  9 14:03:14 2019' <-> 'Fri Aug  9 14:08:14 2019' from TimePeriod 'onDuty'

If you grep 12:35 for instance it seems that segments addition/removal order is not the same. So it has something to deal with API. In my production case I have a distributed architecture so probably another case.

PS: timeperiod in this example:

object TimePeriod "onDuty" {
   // object attributes
   excludes = []
   ranges = {
     monday = "09:00-10:00"
     tuesday = "09:00-10:00"
     wednesday = "09:00-10:00"
     thursday = "12:35-14:05"
     friday = "09:00-10:00"
     saturday = "09:00-10:00"
     sunday = "09:00-10:00"
   }
}

darkweaver87 avatar Aug 08 '19 12:08 darkweaver87

Sounds like #7239 if you always need to restart twice.

dnsmichi avatar Aug 08 '19 12:08 dnsmichi

You're right, the double restart did the trick but in a distributed architecture it seems that is_inside is always to true. I tried to restart both masters twice but it doesn't work :-/

darkweaver87 avatar Aug 08 '19 14:08 darkweaver87

We're a bit busy with RC testing, so debugging this may take a while. Writing from my ipad here. Meanwhile, I'd suggest to analyse in the debug logs if there's an update between the masters on TimePeriods. I doubt it, but why should it be true all the time. Also, check that the TimePeriod object has paused=false being active. Not sure why that would influence segment calculation, but that's code I am not familiar with.

If you're brave, use the centos7-dev vagrant box, compile icinga2 and add breakpoints in gdb for these calculations. The development docs have more insights.

dnsmichi avatar Aug 08 '19 16:08 dnsmichi

OK I understand, no problem. I actually have a workaround by defining a timeperiod by manually computing offDuty periods :-) The paused value is true. Here is the complete output of the console:

<1> => DateTime()
{
	type = "DateTime"
	value = 1565673496.336826
}
<2> => get_time_period("offDuty")
{
	__name = "offDuty"
	active = true
	display_name = "offDuty"
	excludes = [ "onDuty" ]
	extensions = {
		DbObject = {
			type = "Object"
		}
	}
	ha_mode = 0.000000
	includes = [ ]
	is_inside = true
	name = "offDuty"
	original_attributes = null
	package = "shinken_migration"
	pause_called = false
	paused = true
	prefer_includes = true
	ranges = {
		friday = "00:00-23:59"
		monday = "00:00-23:59"
		thursday = "00:00-23:59"
		tuesday = "00:00-23:59"
		wednesday = "00:00-23:59"
	}
	resume_called = false
	segments = [ {
		begin = 1565647200.000000
		end = 1565733540.000000
	}, {
		begin = 1565733600.000000
		end = 1565819940.000000
	} ]
	source_location = {
		first_column = 1.000000
		first_line = 93.000000
		last_column = 27.000000
		last_line = 93.000000
		path = "/var/lib/icinga2/api/packages/shinken_migration/35629376-2ba1-476b-8dd6-173084a7a540/zones.d/global-templates/timeperiods.conf"
	}
	start_called = true
	state_loaded = true
	stop_called = false
	templates = [ "offDuty", "legacy-timeperiod" ]
	type = "TimePeriod"
	update = {
		arguments = [ "tp", "begin", "end" ]
		deprecated = false
		name = "Internal#LegacyTimePeriod"
		side_effect_free = false
		type = "Function"
	}
	valid_begin = 1565647200.000000
	valid_end = 1565819940.000000
	vars = null
	version = 0.000000
	zone = "global-templates"
}
<3> => get_time_period("onDuty")
{
	__name = "onDuty"
	active = true
	display_name = "onDuty"
	excludes = [ ]
	extensions = {
		DbObject = {
			type = "Object"
		}
	}
	ha_mode = 0.000000
	includes = [ ]
	is_inside = true
	name = "onDuty"
	original_attributes = null
	package = "shinken_migration"
	pause_called = false
	paused = true
	prefer_includes = true
	ranges = {
		"2018-05-10" = "00:00-23:59"
		"2019-01-01" = "00:00-23:59"
		"2019-05-01" = "00:00-23:59"
		"2019-05-08" = "00:00-23:59"
		"2019-05-30" = "00:00-23:59"
		"2019-07-14" = "00:00-23:59"
		"2019-08-15" = "00:00-23:59"
		"2019-11-01" = "00:00-23:59"
		"2019-11-11" = "00:00-23:59"
		"2019-12-25" = "00:00-23:59"
		"2020-01-01" = "00:00-23:59"
		"2020-04-13" = "00:00-23:59"
		"2020-05-01" = "00:00-23:59"
		"2020-05-08" = "00:00-23:59"
		"2020-05-21" = "00:00-23:59"
		"2020-07-14" = "00:00-23:59"
		"2020-08-15" = "00:00-23:59"
		"2020-11-01" = "00:00-23:59"
		"2020-11-11" = "00:00-23:59"
		"2020-12-25" = "00:00-23:59"
		"2021-01-01" = "00:00-23:59"
		"2021-04-05" = "00:00-23:59"
		"2021-05-01" = "00:00-23:59"
		"2021-05-08" = "00:00-23:59"
		"2021-05-13" = "00:00-23:59"
		"2021-07-14" = "00:00-23:59"
		"2021-08-15" = "00:00-23:59"
		"2021-11-01" = "00:00-23:59"
		"2021-11-11" = "00:00-23:59"
		"2021-12-25" = "00:00-23:59"
		"2022-04-18" = "00:00-23:59"
		"2022-05-26" = "00:00-23:59"
		"2023-04-10" = "00:00-23:59"
		"2023-05-18" = "00:00-23:59"
		"2024-04-01" = "00:00-23:59"
		"2024-05-09" = "00:00-23:59"
		"2025-04-21" = "00:00-23:59"
		"2025-05-29" = "00:00-23:59"
		"2026-04-06" = "00:00-23:59"
		"2026-05-14" = "00:00-23:59"
		"2027-03-29" = "00:00-23:59"
		"2027-05-06" = "00:00-23:59"
		"2028-04-17" = "00:00-23:59"
		"2028-05-25" = "00:00-23:59"
		"2029-04-02" = "00:00-23:59"
		"2029-05-10" = "00:00-23:59"
		"2030-04-22" = "00:00-23:59"
		"2030-05-30" = "00:00-23:59"
		"2031-04-14" = "00:00-23:59"
		"2031-05-22" = "00:00-23:59"
		"august 15" = "00:00-23:59"
		"december 25" = "00:00-23:59"
		friday = "00:00-09:59,18:01-23:59"
		"january 1" = "00:00-23:59"
		"july 14" = "00:00-23:59"
		"may 1" = "00:00-23:59"
		"may 8" = "00:00-23:59"
		monday = "00:00-09:59,18:01-23:59"
		"november 1" = "00:00-23:59"
		"november 11" = "00:00-23:59"
		saturday = "00:00-23:59"
		sunday = "00:00-23:59"
		thursday = "00:00-09:59,18:01-23:59"
		tuesday = "00:00-09:59,18:01-23:59"
		wednesday = "00:00-09:59,18:01-23:59"
	}
	resume_called = false
	segments = [ {
		begin = 1565647200.000000
		end = 1565683140.000000
	}, {
		begin = 1565712060.000000
		end = 1565733540.000000
	}, {
		begin = 1565733600.000000
		end = 1565769540.000000
	}, {
		begin = 1565798460.000000
		end = 1565819940.000000
	} ]
	source_location = {
		first_column = 1.000000
		first_line = 116.000000
		last_column = 26.000000
		last_line = 116.000000
		path = "/var/lib/icinga2/api/packages/shinken_migration/35629376-2ba1-476b-8dd6-173084a7a540/zones.d/global-templates/timeperiods.conf"
	}
	start_called = true
	state_loaded = true
	stop_called = false
	templates = [ "onDuty", "legacy-timeperiod" ]
	type = "TimePeriod"
	update = {
		arguments = [ "tp", "begin", "end" ]
		deprecated = false
		name = "Internal#LegacyTimePeriod"
		side_effect_free = false
		type = "Function"
	}
	valid_begin = 1565647200.000000
	valid_end = 1565819940.000000
	vars = null
	version = 0.000000
	zone = "global-templates"
}

Unfortunately I won't have the time to debug this for the moment :-/

darkweaver87 avatar Aug 13 '19 05:08 darkweaver87

It might be related to serializer fixes with the segments, discussed and fixed with @Elias481 earlier. Though I'm in the midst of 2.11 testing and Icinga meetup Linz preparations, so it will take a while.

dnsmichi avatar Aug 13 '19 06:08 dnsmichi

OK no problem I understand :-)

darkweaver87 avatar Aug 20 '19 07:08 darkweaver87

I don't think is related to serializer fixes in this case, at least I don't see any connection. And I have no idea how it could be in "start_called" state without beeing intiaielized and have started a time to refresh them. Also I do not exactly see what paused mean (but as far as I see it cannot have the effect that start_called is the state but timeranges not initializes). I don't have a HA setup for testing in place and no time currenlty.

Elias481 avatar Aug 24 '19 20:08 Elias481

@darkweaver87 Please could you test whether the snapshot packages are still affected. If yes, please also test the packages from here – pick your OS from the "binary" column and then "Download" the "Job artifacts".

Al2Klimov avatar Feb 27 '20 09:02 Al2Klimov

@Al2Klimov: OK I will do but I won't have the time in the next 2 months (at least) unfortunatelly. May it be better to include this in a unit test suite ?

darkweaver87 avatar Feb 27 '20 12:02 darkweaver87

Don't worry, we have time. 🙂

Al2Klimov avatar Feb 27 '20 12:02 Al2Klimov

OK anyway thanks for trying to fix it :-)

darkweaver87 avatar Feb 27 '20 12:02 darkweaver87

If the "Job artifacts" are already gone while you're going to test them, please let me know – I'll re-create them.

Al2Klimov avatar Feb 27 '20 13:02 Al2Klimov

PING @darkweaver87

Al2Klimov avatar Aug 10 '21 19:08 Al2Klimov

Hello,

Sorry but I don't work for my previous company anymore, thus I don't have a working environment to validate/invalidate this.

Rémi

darkweaver87 avatar Oct 19 '21 06:10 darkweaver87

@darkweaver87 Please could you test whether the snapshot packages are still affected. If yes, please also test the packages from here – pick your OS from the "binary" column and then "Download" the "Job artifacts".

I am experiencing the same issue in icinga2 2.10.3-2+deb10u1 (Debian Buster). Could I verify the fix?

jinowolski avatar May 17 '22 09:05 jinowolski

Yes, with these packages:

https://git.icinga.com/packaging/deb-icinga2/-/jobs/153741/artifacts/download

Al2Klimov avatar May 19 '22 08:05 Al2Klimov

ref/IP/44756

Al2Klimov avatar Feb 22 '23 14:02 Al2Klimov