unsubscribe_event fails in delete_device
|
|
|
|---|---|
|
Hi, In a high level DS I subscribe to multiple events/low level DS. I call subscribe_event in init_device and unsubscribe_event in delete_device. In init_device: eventid = self._device.subscribe_event("XXX", PyTango.EventType.CHANGE_EVENT, self._callback, stateless=False) self._eventids.append(eventid) eventid = self._device.subscribe_event("YYY", PyTango.EventType.CHANGE_EVENT, self._callback, stateless=False) self._eventids.append(eventid) In delete_device: for eventid in self._eventids: self._device.unsubscribe_event(eventid) The fellowing error appears everytimes during delete_device execution: DevError[ desc = Failed to disconnect from event! Error while trying to unsubscribe the heartbeat ZMQ socket from the channel heartbeat publisher ZMQ message: No such file or directory origin = ZmqEventConsumer::disconnect_event reason = API_ZmqFailed severity = ERR] ]' After several restarts the fellowing error appears during init_device execution: DevError[ desc = Failed to delay event! Error while communicating with the ZMQ main thread ZMQ message: Connection refused origin = DelayEvent::DelayEvent reason = API_ZmqFailed severity = ERR] ] Im using PyTango 8.1.6, Tango 8.1.2c Best regards, Philippe |
|
|
|
|---|---|
|
|
Hi Philippe, We've been using similar devices for a while and your code looks fine, so I assume it is a configuration issue. My first guess would be something about the tango host. ZMQ events can be a bit picky about it since it includes the host name inside the channel name. For instance, I had similar problems when I used an IP address as a tango host. Also, is your callback called when the device is running? Do you get the 'Event channel not responding anymore' DevError? Some more information about your config could help as well (OS and zeromq version). |
|
|
|
|---|---|
|
|
Hi Vincent, Thanks a lot for a so quick answer. First of all, here is the environment I use : Windows 7 PyTango 8.1.6 Tango 8.1.2b libzmq 4.0.4 pyzmq 14.4.1 My TANGO_HOST was effectively 127.0.0.1. I changed to localhost, it is the same problem. Here is the code I use to test the execution of delete_device : import PyTango dev = PyTango.DeviceProxy("ZZZ") adm = PyTango.DeviceProxy(dev.adm_name()) adm.Kill() When the High level DS is running and I stop a Low level DS, the callback associated with this DS report "'Event channel not responding anymore' DevError" as you mentionned. When I restart the Low level DS, the callback is called again, that's fine ! Best Regards Philippe |
|
|
|
|---|---|
|
|
Hi Philippe, Sorry for the late answer. If I understand correctly, your event channel works fine until your try to unsubscribe, is that it? There are other (and easier) ways to test your delete_device method: - Run the INIT command: it will successively call delete_device and init_device on your device object - Send a keyboard interruption to the device server Let me know if you get different results using those approaches. Cheers |
|
|
|
|---|---|
|
I'm having a similar problem here, with PyTango 8.1.8. The problem occurs only when a server is running more than one device. The INIT command seems to work properly in both the devices that are running on the same server, but if I kill the process with a keyboard interruption, then I get theFailed to disconnect from event! error. In this case I have to press CTRL+C several times, and at the end I get a segmentation fault here: #0 0x00007f6867101b98 in set_svr_shutting_down (this=0x1db1690) at ../../../../lib/cpp/server/utils.h:893 In case of a single device server, I don't have any problem. Please note that in both cases the events work properly: I have problems only when trying to kill the multi-device servers. |