Karabo-bridge-glimpse ZMQ error when using bash

[skuschel]

I am currently using the karabo-bridge-glimpse at SQS to dignose errors in the datastream. Whenever I use it in the default shell it works. However, I like to work in the bash and when I start the bridge in from the bash I get a ZMQ error. Here is how to reporduce it:

1) in default shell (this works as expected). After logging onto exflonc35:

module load anaconda3/5.2
karabo-bridge-glimpse tcp:10.253.142:6666

2) does not work. After looging onto exflonc35:

bash
module load anaconda3/5.2
karabo-bridge-glimpse tcp:10.253.142:6666

This yield the following error message:

Traceback (most recent call last):
  File "/software/anaconda3/5.2/bin/karabo-bridge-glimpse", line 11, in <module>
    sys.exit(main())
  File "/software/anaconda3/5.2/lib/python3.6/site-packages/karabo_bridge/cli/glimpse.py", line 184, in main
    client = Client(args.endpoint)
  File "/software/anaconda3/5.2/lib/python3.6/site-packages/karabo_bridge/client.py", line 60, in __init__
    self._socket.connect(endpoint)
  File "zmq/backend/cython/socket.pyx", line 580, in zmq.backend.cython.socket.Socket.connect
  File "zmq/backend/cython/checkrc.pxd", line 25, in zmq.backend.cython.checkrc._check_rc
zmq.error.ZMQError: Protocol not supported

I did not create my own python installation, and whereis karabo-bridge-glimpse points to the same executable in both cases.

Its probably a low priority, but I wanted to make you aware that there is something in the environment. Strangely, my own python scripts, which are using the karabo-bridge.Client to connect to the live stream work in both shells.

Cheers Stephan

[takluyver]

Normally the endpoint specification has a // in, like tcp://10.253.142:6666. Does that make a difference?

[skuschel]

That was a typo, I actually used tcp://10.253.142:6666 in both instances.

However, I just double checked my configuration and figured that I a old .bashrc, as the live servers are using a different home than the maxwell server and I updated it in only in one place. Sorry for bothering you and thank you very much for your fast response!