forked from M-Labs/artiq
f1fd42ea98
This partially reverts commit b5e1bd3fa2,
which had removed keepalive. This, however, led to experiments
hanging forever if the core device had dropped the connection
(e.g. to a kernel CPU panic, or the device being rebooted).
The chosen keepalive settings are fairly conservative (with the
10 s timeout) to avoid any possible interaction with smoltcp's
3 s ARP try interval (see GitHub issue #1150), even though this
should be a non-issue now due to the larger ARP cache.
29 lines
1.1 KiB
Python
29 lines
1.1 KiB
Python
import sys
|
|
import socket
|
|
import logging
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
def set_keepalive(sock, after_idle, interval, max_fails):
|
|
if sys.platform.startswith("linux"):
|
|
sock.setsockopt(socket.SOL_SOCKET, socket.SO_KEEPALIVE, 1)
|
|
sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_KEEPIDLE, after_idle)
|
|
sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_KEEPINTVL, interval)
|
|
sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_KEEPCNT, max_fails)
|
|
elif sys.platform.startswith("win") or sys.platform.startswith("cygwin"):
|
|
# setting max_fails is not supported, typically ends up being 5 or 10
|
|
# depending on Windows version
|
|
sock.ioctl(socket.SIO_KEEPALIVE_VALS,
|
|
(1, after_idle * 1000, interval * 1000))
|
|
else:
|
|
logger.warning("TCP keepalive not supported on platform '%s', ignored",
|
|
sys.platform)
|
|
|
|
|
|
def initialize_connection(host, port):
|
|
sock = socket.create_connection((host, port))
|
|
set_keepalive(sock, 10, 10, 3)
|
|
logger.debug("connected to %s:%d", host, port)
|
|
return sock
|