Hi there. I just got a strange issue which i'm going to describe. It's been a while that i have setup a private network with 3 machines. Everything always worked fine until this strange issue. Let's name the three machines A, B, C. Suddenly this happens:
A can ping C (and viceversa)
B can ping C (and viceversa)
A can't ping B (and viceversa)
All three machines are running Windows 10 1909. I tried to reboot both A and B, restart the zerotier services, change and delete A and B ips in the controller, disable and re-enable A and B in the controller. Nothing of those worked. I managed to solve this issue following those steps:
What could be happened? Is there anything else to check before do the uninstall/reinstall trick?
Thank you!
same here, reboot solve the issue for me.
same here, https://zerotier.atlassian.net/servicedesk/customer/portal/1/ZT-2212?created=true
Hello Zerotier Support,
I cannot connect to my zt1 client these days, but i can connect another zt1 client in the same remote network.
using command 'zerotier-cli peers' shows the zt1 client disconnected from all the PLANET servers.
but i tried to re-generate (remove /var/lib/zerotier-one) zt1 ID for the client, and then using new ID to join the same network, it works again.
so i think may the zt1 ID was blocked by PLANET server?
the old zt1 client ID: 4efc (not working)
the new zt1 client ID: ce9f (it works)both new and old ID joins network 93***9e
now i using new ID without any problem,
i just want know what reason cause that problem?thanks,
Pekaikon NorckonComponent/s or products
ZeroTier One
I have the same issue with 2 computers running Linux and Zerotier 1.4.16.
However the drops I have will only for a few minutes/hours and it will comeback without any action.
Today it happens again. From my research it seems that it can't ping some machines when those machines have many PLANET peers with the "RELAY" status. When i can get the "DIRECT" status everything seems to work good. Now the question is: is there any way to connect again without changing any configuration for trying to make the DIRECT peers connection works?
Today happened again, i can't ping one machine but everything is online. This is very annoying sadly...
@graphixillusion @jbaptisteblot @LiyuanLucasLiu @dfc643 I'm going over our GitHub issues in preparation for our pending 1.6.0 release; that should hopefully take care of a few things. If you're still having this issue ATM, can you please check the following & let us know otherwise so we can try to catch it? Thanks!
@unquietwiki For now the only way to mitigate this is stop the services for about 10 mins and restart it: this problem is different from the ones you linked in the other forums (check #1214). Let's see with the new 1.6.0 version when it will comes out.
@graphixillusion Acknowledged. I'll make the rounds again when that's ready (days/weeks on that). Thank you for your patience.
@graphixillusion sorry if we've asked you before, but what kind of router is that windows machine behind? just curious.
We've seen the multiply paths thing happen, but not to where it stops communication.
some work-around guesses:
peers.d and local.conf go in "ProgramDataZeroTierOne"
@laduke my router is a ISP branded one, maded by Sercom. As i said, for now the workaround i'm currently using is to stop the zerotier's service for about 10 min and restart it without touching anything else. Next time i will try your first option and see if it will works: stop zerotier, delete the peers.d folder and restart zerotier with no wait time. In the meantime, could you post a valid local.conf config which should fix this issue? Thank you.
@graphixillusion https://discuss.zerotier.com/t/asymmetric-routing-problem-with-site-to-site/176 this showed up on our discussion board today. Might be related?
@unquietwiki nope, it seems a different issue. In my case, the zerotier client failed to connect to the planets after my internet goes down: when internet is back, the only way to make it works again is do what i already described (stop zerotier service for 10 mins and then restart it)
@graphixillusion Thanks; wanted to see if that was a possibility.
@laduke Today i tried the workaround you suggested:
Nothing: doesn't work. Currently the only way that it works everytime is to stop the service for about 10 min, then restart it.
Thanks for trying.
here's the local.conf
{"settings": {"portMappingEnabled": false}}
This may not be suitable for your environment if it's relying on UPnP.
Does rebooting the router work around the issue? Don't worry about testing this if it's disruptive!
@laduke nope, reboot the router doesn't fix the problem.
I'm testing the newly 1.6.1. I will report if something is changed.
Just want to report that the problem is still present in the last 1.6.2 version.