Those are background threads of libzmq, as mentioned here: https://github.com/zeromq/libzmq/blob/master/NEWS
Jupyter uses ZMQ to connect to the running kernels on the local machine. There's not much to be done for the threads besides passing messages, so they won't accumulate a lot of CPU time. Restarting the kernel(s) will close the ZMQ connections and open new ones, maybe that has an effect on the background threads and/or their memory consumption? That was suggested here: https://github.com/spyder-ide/spyder/issues/8766 Don't know why these threads should consume much memory... maybe the kernel or Jupyter are unresponsive, and ZMQ messages get buffered beyond reasonable limits? Jupyter does act as a bridge between the kernels (via ZMQ) and frontends (via Websockets). I think some message buffering was added months ago for cases where the websocket connection is lost and gets re-established later, so the browser can still retrieve the cell output. But I wouldn't expect cell output to be big enough to cause OOM problems. Also, the Spyder issue linked above mentions there was no significant kernel activity. Hope that helps, Roland -- You received this message because you are subscribed to the Google Groups "Project Jupyter" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/jupyter/df2cb4dd-559d-45ed-9c58-911e9bbcf168%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
