Closed camilomarino closed 4 years ago
Thanks for raising this issue. This is a known "feature" of ZMQ, @camilomarino. It was actually imageZMQ issue: #1! "Hangs at...". The most elegant fix is a great one from Pat Ryan @youngsoul https://github.com/jeffbass/imagezmq/issues/1#issuecomment-489476219. Take a look at that that. It may work well for you.
In my own projects, I use a "with Patience()
" try / except block like this one in my imagenode main loop: see imagenode.py (see lines 33-46).
The "hanging" of the REQ client when the REP server restarts is a known ZMQ "feature" and is there by design (so that a dropped REQ won't go unnoticed by the sender). I use this ZMQ feature as a part of my own yin-yang-ranch project design -- I want the RPi's to know if they need to deal with a non-responsive imagehub.
Please the FAQ page in the documentation. It has a more complete discussion of this. See FAQ Question "Why does image sender need to restart when the image hub program restarts?"
Hello, when the server is turned off, the client cannot reconnect. I wanted to know if there is any way to tell the client that in case of not receiving the reply message (for example in 5 seconds) and try to reconnect.