@future-ram-24559 there's a netclient installer that...I think?...is notarized. Probably needs to be tested. worked on my machine but installed / uninstalled several times so not sure if that affected it. https://fileserver.netmaker.org/v0.14.0/netclient-install.pkg

thanks @jolly-london-20127 - Discord tells me the download is potentially dangerous! I can only hope it is right, I'll revert back in a couple of hours, hope you're sleeping right now...

ok here are my findings- prep- I upgraded the Server to 0.14.0 and deleted my local machine from the UI 1. Thanks so much for doing this, it should make my life so much easier (soon) 2. installer is validly signed, notarised and put together. 3. There is still a gatekeeper warning in the install logs but that may be ok 4. the binaries seem to be installed in the correct place and they show all green ticks for signing and security 5. I see there's still some removal code in the installer- thanks I was putting stuff all over the place but I now have an uninstall script 6. Join with token- didn't work, got this error (network name is 'infra_access') 'unable to authenticate: Post "https:///api/nodes/adm/infra_access/authenticate": http: no Host in request URL ' created a new network called 'test' in 10.192.0.0/24 and issued the join command.... [netclient] 2022-05-17 17:59:39 joining test at api.wg.domain.com.au:443 [netclient] 2022-05-17 17:59:39 starting wireguard [netclient] 2022-05-17 17:59:52 certificates/key saved [netclient] 2022-05-17 18:00:22 unable to connect to broker, retrying ... [netclient] 2022-05-17 18:00:22 could not connect to broker broker.wg.domain.com.au connect timeout [netclient] 2022-05-17 18:00:22 connection issue detected.. attempt connection with new certs [netclient] 2022-05-17 18:00:22 certificates/key saved [netclient] 2022-05-17 18:00:54 could not connect to broker at broker.wg.domain.com.au:8883 [netclient] 2022-05-17 18:00:54 sent a node update to server for node iMaclocal , xxxxxxx-67a0-4x75-ax90-xxxxxxxxxxx But this is probably an issue at my end- Traefik is supposed to create broker.wg.domain.com.au The node is listed in 'test' network as healthy, will check again in 10 So I think that is success!

Yeah these are upgrade issues. the token issue is likely a leftover token from before upgrade. If you generate a new token it should work. And then the next issue is with the broker connection meaning either DNS is incorrect or 8883 is not reachable. - gatekeeper warning - not sure what this is, any idea what I need to do to avoid this?

Um... just finished ripping out and replacing and playing a bit with 0.14.0 ... Noticed on macs (monterey and big sur) that I cant find the netclient process anymore, and /etc/hosts isnt updated. I uninstalled 0.12.0 and reinstalled with 0.14.0 in all cases (Linux and Mac) Am I having a problem or is this by design?

let's move this to a thread

Ok

ok sorry for the time taken to reply- here is the latest. I created a DNS record for broker.xxx.com verified that it's working and verified that port 8883 is open on the host. Netclient seems to be working fine, but still won't connect. here's the latest Netclient pull- [netclient] 2022-05-18 11:47:20 initializing network test [netclient] 2022-05-18 11:47:21 started daemon for server broker.wg.domain.com.au [netclient] 2022-05-18 11:47:21 netclient daemon started for server: broker.wg.domain.com.au [netclient] 2022-05-18 11:47:52 unable to connect to broker, retrying ... [netclient] 2022-05-18 11:47:52 could not connect to broker broker.wg.domain.com.au connect timeout [netclient] 2022-05-18 11:47:52 connection issue detected.. attempt connection with new certs [netclient] 2022-05-18 11:47:52 register at https://api.wg.domain.com.au:443/api/server/register [netclient] 2022-05-18 11:47:52 certificates/key saved [netclient] 2022-05-18 11:47:53 shutting down netclient daemon [netclient] 2022-05-18 11:47:53 checkin routine closed [netclient] 2022-05-18 11:47:53 shutdown complete So it looks like a setup issue at the Server end. I'll have a look at the logs there

have you checked out the troubleshooting gist? https://gist.github.com/mattkasun/face2a7c1f32031a2126ff7243caad12

might be making some progress, your install docs say 'Open ports 443, 80, 53, 8883, and 51821-51830/udp on the VM firewall and in cloud security settings.' But my VM is not answering on port 53, looking at this now

port 53 would not cause this issue as far as I know

ok I just went back to the official docker-compose and it doesn't mention port 53, so I'll go back to the troubleshooting gist you provided

main things to check are: 1. MQ logs 2. is DNS (broker.domain) pointing to the VM 3. is 8883 open and reachable

looks like I've mangled the mq conf, please hold...

ok I think we are good- sudo netclient pull Password: [netclient] 2022-05-18 12:53:20 No network selected. Running Pull for all networks. [netclient] 2022-05-18 12:53:26 certificates/key saved

wohoo!

sorry, spoke too soon, there's no errors on the Server, but the client is still having a timeout connecting to broker.domain. I'll see what I can figure out with mq, just a sec

verify -verbose -CAfile /root/netmaker/certs/root.pem /root/netmaker/certs/server.pem /root/netmaker/certs/server.pem: OK

on the client- openssl verify -verbose -CAfile /etc/netclient/broker.wg.domain.com.au/root.pem /etc/netclient/broker.wg.domain.com.au/client.pem Error loading file /etc/netclient/broker.wg.domain.com.au/root.pem

I'll try it on a new machine- /etc/netclient aliases to /private/etc/netclient and seems to be full of old netclient stuff of mine

So the certificates are ok, mq logs are ok, but you cant connect to the broker

does nslookup for broker.domain resolve to the correct endpoint?

yes, the DNS resolves correctly and the port is open, checked with online port checker. Installing on a freshly erased Apple Silicon machine, I got the following- [netclient] 2022-05-19 20:09:38 joining test at api.wg.domain.com.au:443 [netclient] 2022-05-19 20:09:38 starting wireguard [netclient] 2022-05-19 20:09:45 certificates/key saved [netclient] 2022-05-19 20:10:15 unable to connect to broker, retrying ... [netclient] 2022-05-19 20:10:15 could not connect to broker broker.wg.domain.com.au connect timeout [netclient] 2022-05-19 20:10:15 connection issue detected.. attempt connection with new certs [netclient] 2022-05-19 20:10:15 certificates/key saved [netclient] 2022-05-19 20:10:46 could not connect to broker at broker.wg.domain.com.au:8883 [netclient] 2022-05-19 20:10:46 sent a node update to server for node Servicemax-2local , axxxce5-30a1-40ax0-877a-xxx So the client does connect and register (MUCH slower than it used to be, in case that's important), and on this machine there is no weird aliasing happening to /etc/netclient, so that bit seems to be a non-issue. But something funny about the mq broker connection. I've sent a copy of the logs for the 'join' command via email. Let me know if you have any ideas...

have you tried all the MQ troubleshooting https://gist.github.com/mattkasun/face2a7c1f32031a2126ff7243caad12

Hi... how do you restart netclient on a mac? I cant find it anymore when I list launchtl processes

launchd unload & launchd load?

launchctl*

should be com.gravitl.netclient

Weird, I can find it after rebooting now (launchctl list)

Thanks