The network connection is large, causing the to run full when the salt detection is alive.

[menglong81]

Reading the /proc/net/tcp file for a machine with a large network connection will consume CPU. So we changed to use the ss command instead of the netstat command to get the network connection first. The ss uses the tcp_diag in the TCP protocol stack so it will be faster than Netstat command is much faster

[menglong81]

solve issue:
https://github.com/menglong81/salt/tree/develop

[WalkerWang731]

In fact currently socket.py module does not support the NETLINK_INET_DIAG protocol, so it can only be replaced with the ss tool. When the socket.py module supports more NETLINK protocols, you can using socket.py module to implement

[Ch3LL]

just to clarify are you suggesting we migrate from netstat to ss in the socket.py module?

[WalkerWang731]

I mean, python's official socket library not support NETLINK_INET_DIAG. so can be use ss first. if salt use ss to replace netstat, maybe is nice try. however old release of linux maybe unsupport, so ss before netstat

[menglong81]

master_alive_interval options
Salt will call the netstat command when it performs health check. However, if the number of system connections is large, it will consume cpu because it reads the file /proc/net/tcp
So we modified the code and hope to adopt it

[Ch3LL]

thanks for clarifying and thanks for teh PR!

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

If this issue is closed prematurely, please leave a comment and we will gladly reopen the issue.

[stale[bot]]

Thank you for updating this issue. It is no longer marked as stale.