Ever-increasing worker goroutines

[4nte]

My volantmq broker eventually stops working due to too many workers.
I have unstable clients that frequently reconnect due to flaky network connections.

What is the best way to mitigate this?

goroutine profile: total 404
365 @ 0x63f430 0x614108 0x6140de 0x613e0b 0xa2c7cf 0x66a611
#	0xa2c7ce	github.com/VolantMQ/volantmq/types.(*pool).worker+0x9e	/go/src/github.com/VolantMQ/volantmq/types/pool.go:103

23 @ 0x63f430 0x63952a 0x638af5 0x6d29d5 0x6d394f 0x6d3931 0x7a3acf 0x7b7138 0xa40bd5 0x764403 0x76456f 0xa4c84a 0xa4c13e 0x66a611
#	0x638af4	internal/poll.runtime_pollWait+0x54					/usr/local/go/src/runtime/netpoll.go:184
#	0x6d29d4	internal/poll.(*pollDesc).wait+0x44					/usr/local/go/src/internal/poll/fd_poll_runtime.go:87
#	0x6d394e	internal/poll.(*pollDesc).waitRead+0x1ce				/usr/local/go/src/internal/poll/fd_poll_runtime.go:92
#	0x6d3930	internal/poll.(*FD).Read+0x1b0						/usr/local/go/src/internal/poll/fd_unix.go:169
#	0x7a3ace	net.(*netFD).Read+0x4e							/usr/local/go/src/net/fd_unix.go:202
#	0x7b7137	net.(*conn).Read+0x67							/usr/local/go/src/net/net.go:184
#	0xa40bd4	github.com/VolantMQ/volantmq/transport.(*conn).Read+0x54		/go/src/github.com/VolantMQ/volantmq/transport/conn.go:41
#	0x764402	bufio.(*Reader).fill+0x102						/usr/local/go/src/bufio/bufio.go:100
#	0x76456e	bufio.(*Reader).Peek+0x4e						/usr/local/go/src/bufio/bufio.go:138
#	0xa4c849	github.com/VolantMQ/volantmq/connection.(*reader).readPacket+0x79	/go/src/github.com/VolantMQ/volantmq/connection/reader.go:129
#	0xa4c13d	github.com/VolantMQ/volantmq/connection.(*reader).routine+0x11d		/go/src/github.com/VolantMQ/volantmq/connection/reader.go:80

[troian]

thanks, i'll give it a think

[4nte]

@troian have you figured anything out? Any general pointers would be appreciated (if you have any, of course).
I'll try to open a PR in the following days.