Open daneos opened 6 years ago
It looks like any macro manipulating Pool elements is affected. This points to the event system as the source of the problem, but I guess in this case errors should be logged even with the default log level. I'll test this with a higher log level and post my results.
door/bl04bm/devel [2]: lsctrl
Name Type Class Module
------------------------------ -------------- ----------------------- -----------------
bl-04bmcab03-ctl-ipap01 Motor SolarisIcepapCtrl solaris_motor
bl-04bmcab04-ctl-ipap01-BARE Motor IcepapController IcePAPCtrl
tango-r1 CTExpChannel TangoAttrCTController TangoAttrCTCtrl
test CTExpChannel TangoAttrCTController TangoAttrCTCtrl
door/bl04bm/devel [3]: udefctrl test
door/bl04bm/devel [4]: lsctrl
Name Type Class Module
------------------------------ -------------- ----------------------- -----------------
bl-04bmcab03-ctl-ipap01 Motor SolarisIcepapCtrl solaris_motor
bl-04bmcab04-ctl-ipap01-BARE Motor IcepapController IcePAPCtrl
tango-r1 CTExpChannel TangoAttrCTController TangoAttrCTCtrl
test CTExpChannel TangoAttrCTController TangoAttrCTCtrl
door/bl04bm/devel [5]: defmeas testmg01 SR
Created None
Sorry for the late reply, I was out-of-office.
I agree, it looks like a problem with the events. Could you please try the following instructions: https://github.com/taurus-org/taurus/issues/658#issuecomment-369616943 and report back the results.
Also, please attach the logs of the Pool and the MacroServer started with the --log-level=debug
.
When creating controllers or elements the output is
Created None
and the element is not visible until MacroServer restart. I can see Tango devices are created, but MacroServer isn't aware of that. Looks like some problem with communication between Pool and MacroServer.I think I saw this problem before, but couldn't find it in the issues.
Now for the hard part: there is absolutely no useful information in Pool or MacroServer logs (at least when using default log level). I tried to reproduce this problem on some other VM, and that's where I got totally lost, because it works there, despite all VMs are set up using the same Ansible configuration.
I guess it has something to do with the configuration of this particular VM, but I don't know where to start looking.