- **fix deadlock in ospf route introducd by #958 **
- **use random peer id for foreign network entry, because ospf route algo need peer id change after peer info version reset. this may interfere route propagation and cause node residual**
- **allow multiple nodes broadcast same network ranges for subnet proxy**
- **bump version to v2.3.2**
sometimes route table may not be updated in time, so some dead nodes are still showing in the peer list.
when generating ipv4-peer table, we should avoid these dead devices overrides the entry of healthy nodes.
this may fix following problem:
1. cpu 100% when large number of nodes in network.
2. high cpu usage when large number of foreign networks.
3. packet loss when new node enters/exits.
4. old routes not cleand and show as an obloleted entry.
Some devices have ipv6 but don't allow input connection, this patch add hole punching for these devices.
- **add v6 hole punch msg to udp tunnel**
- **send hole punch packet when do ipv6 direct connect**
1. dns resolver should be global unique so dns cache can work. avoid dns query influence hole punching.
2. when system dns failed, fallback to hickory dns.
also some improvements:
1. add magic dns option in gui.
2. allow icmp proxy fail on android
3. when no_tun is enabled, android do not start vpn service
Co-authored-by: Your Name <you@example.com>
This patch implements:
1. A dns server that handles .et.net. zone in local and forward all other queries to system dns server.
2. A dns server instance which is a singleton in one machine, using one specific tcp port to be exclusive with each other. this instance is responsible for config system dns and run the dns server to handle dns queries.
3. A dns client instance that all easytier instance will run one, this instance will try to connect to dns server instance, and update the dns record in the dns server instance.
this pr only implements the system config for windows. linux & mac will do later.
* optimize memory issues
1. introduce jemalloc support, which can dump current memory usage
2. reduce the GlobalEvent broadcaster memory usage.
3. reduce tcp & udp tunnel memory usage
TODO: if peer conn tunnel hangs, the unbounded channel of peer rpc
may consume lots of memory, which should be improved.
* select a port from 15888+ when port is 0
* support ipv6 stun
* show interface and public ip in cli node info
* direct conn should keep trying unless already direct connected
* peer should use conn with smallest latency
* deprecate ipv6_listener, use -l instead
