Logspout fails to "spout" after TCP/UDP connection breaks and doesn't try to reconnect

[Hossy]

I'm using Logspout with the logstash module. I have tried using UDP and TCP connections with and without the RETRY_STARTUP option from the logstash module. In all cases, if the original connection Logspout established to Logstash is broken, Logspout will never reconnect until the Logspout container is restarted.

When the connection is lost, there is no log activity in the Logspout container. Logspout will only show log activity when a container it is monitoring logs something that needs to be "spouted." When that happens, I see the following logs in the Logspout container.

I've reviewed the following issues and they are either similar but don't solve my problem or didn't provide enough information and it seems they've been abandoned: #330 #377 #386

Note: This is not an issue with docker log rotate like #216 #255 #309 #419 #504 or a large log line like #420.

10/4/2022 6:28:08 PM2022/10/04 18:28:08 pump.Run() event: 21447dd8667c kill
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.Run() event: 21447dd8667c kill
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.Run() event: 21447dd8667c die
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.pumpLogs(): 21447dd8667c stopped
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.Run() event: 98a5cba01a76 network:disconnect
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.Run() event: discovery_co volume:unmount
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.Run() event: rancher-cni volume:unmount
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.pumpLogs(): 21447dd8667c dead
10/4/2022 6:28:18 PM2022/10/04 18:28:18 pump.Run() event: 21447dd8667c stop
10/4/2022 6:28:23 PM2022/10/04 18:28:23 pump.Run() event: 98a5cba01a76 network:connect
10/4/2022 6:28:23 PM2022/10/04 18:28:23 pump.Run() event: discovery_co volume:mount
10/4/2022 6:28:23 PM2022/10/04 18:28:23 pump.Run() event: 21447dd8667c start
10/4/2022 6:28:23 PM2022/10/04 18:28:23 pump.pumpLogs(): 21447dd8667c started, tail: all