CA config is sent to MH every short heartbeat

[JamesUoM]

Currently the CA configuration is sent to MH at the same time as the state ie on the short heartbeat (see scheduler.set_state. The reference MH CA setup has the CA config set to MH every 600s (10mins), from org.opencastproject.capture.impl.ConfigurationManager.properties:

capture.config.remote.polling.interval=600

this implies that the CA config should be sent on the long heartbeat not the short.

The only thing that changes in the CA config is the mindiskspace but I don't think there is any need to but unnecessary strain on the MH admin node by sending this every 10s by default.

[smarquard]

Agreed, it would be good to get this fixed.

[hcaloto]

The polling interval has been set to the long heartbeat, so you can change its value in the config file. Thanks for reporting this.

[smarquard]

The change in 958a89f is not sufficient to set this - it also needs a change in galicaster/scheduler/scheduler.py to remove setconfiguration from set_state and call it separately from do_timers_long

   def set_state(self):
        if self.state.is_error:
            self.ca_status = 'unknown' #See AgentState.java
        elif self.state.is_recording:
            self.ca_status = 'capturing'
        else:
            self.ca_status = 'idle'
        self.logger.info('Set status %s to server', self.ca_status)
        try:
            self.client.setstate(self.ca_status)
            self.client.setconfiguration(self.conf.get_tracks_in_mh_dict())
            self.net = True
            self.emit('net-up')
        except:
            self.logger.warning('Problems to connect to matterhorn server ')
            self.net = False
            self.emit('net-down')
            return

[smarquard]

diff --git a/galicaster/scheduler/scheduler.py b/galicaster/scheduler/scheduler.py
index 39af45b..b7d3a7d 100644
--- a/galicaster/scheduler/scheduler.py
+++ b/galicaster/scheduler/scheduler.py
@@ -69,6 +69,7 @@ class Scheduler(object):

     def do_timers_long(self, sender):
         if self.net:
+            self.client.setconfiguration(self.conf.get_tracks_in_mh_dict())
             self.proccess_ical()
             self.emit('after-process-ical')
         for mp in self.repo.get_next_mediapackages():
@@ -100,7 +101,6 @@ class Scheduler(object):
         self.logger.info('Set status %s to server', self.ca_status)
         try:
             self.client.setstate(self.ca_status)
-            self.client.setconfiguration(self.conf.get_tracks_in_mh_dict())
             self.net = True
             self.emit('net-up')
         except:

[JamesUoM]

+1

On 25/05/15 12:03, Stephen Marquard wrote:

|diff --git a/galicaster/scheduler/scheduler.py b/galicaster/scheduler/scheduler.py
index 39af45b..b7d3a7d 100644
--- a/galicaster/scheduler/scheduler.py
+++ b/galicaster/scheduler/scheduler.py
@@ -69,6 +69,7 @@ class Scheduler(object):

  def do_timers_long(self, sender):
      if self.net:

•        self.client.setconfiguration(self.conf.get_tracks_in_mh_dict())
        self.proccess_ical()
        self.emit('after-process-ical')
    for mp in self.repo.get_next_mediapackages():
  

  @@ -100,7 +101,6 @@ class Scheduler(object):
  self.logger.info('Set status %s to server', self.ca_status)
  try:
  self.client.setstate(self.ca_status)

•        self.client.setconfiguration(self.conf.get_tracks_in_mh_dict())
        self.net = True
        self.emit('net-up')
    except:
  

  |

—
Reply to this email directly or view it on GitHub
#114 (comment).

---

James S. Perrin

Media Technologies Team
J20, Sackville Building
The University of Manchester
Oxford Road, Manchester, M13 9PL

t: +44 (0) 161 275 6945
e: james.perrin@manchester.ac.uk

w: www.manchester.ac.uk/researchcomputing

"The test of intellect is the refusal to belabour the obvious"

- Alfred Bester

[smarquard]

Still an issue in 2.0.x although this code is now in galicaster/opencast/service.py