Still hitting the tailscale drums?
I’m so glad mikrotik decided to stop adding completly new features for now and that they only try to fix bugs with the current features.
Btw, the tailscale binaries are HUGE. Nobody considers this a problem for mikrotik devices?
If MikroTik decides to add TailScale client and its size forces them to make devices with more storage, I say go for it, it will be worth it for just this reason alone. 
you prefer a larger package sob? Its the weekend LOL.
Right size for given purpose is the best. In this case, if it would serve as kind of Trojan horse (although it’s quite unfortunate term, because the meaning related to computers is already taken) to increase storage, it’s the bigger the better.
I think TailScale support should be added to integrate with existing TS networks.
TS is gaining significant momentum as the good alternative to VPN and one very very useful use case is VPN to cloud VPCs.
In my tech bubble, I haven’t seen ZT being used at all actually.
Imagine working on a high-value IoT project with cloud resources and tons of IoT devices; it might be a valuable use case to be able to very easily create a secure network spanning cloud+IoT; emphasis is on easy and on secure.
I think it shouldn’t be added.
And you can run it in a container, can you not?
We have heard that about most VPN technologies by now. It seems there is always one more VPN that MikroTik REALLY have to integrate.
And once they have done so, and a few months pass, the second wave of requests comes in, to have the version updated because it is more and more lacking in features.
I have to agree with the above: when external parties want their VPN to be supported, let them deliver it as a container (as small as possible), and update that container at the rate they desire.
It should be possible in RouterOS Real Soon Now. Active work is being done on containers to facilitate use by VPN services (TUN device).
Hopefully that can finally end the long topics about “Please support YetAnotherNewfangledVPN” and the hundreds of +1 replies.
+1 for that post
My point is that they adopted zerotier but it has limited applicability, not all MT devices can run it.
If tailscale can run on more devices, then it should be adopted if they are relatively equal otherwise.
Let the user decide which package they want to load!
There is an active Tik user [cannot remember his name] that is building a TailScale Container for Tik Devices … however that is strictly limited to ARM based Tiks so currently very limited in scope. As I have stated before ZeroTier is IMO a waste of time and a very poor performer plus as you stated not all MT devices can run it.. ABSOLUTLY there is no question that TailScale is a very good WireGuard Management system.
What’s wrong with the official container? https://hub.docker.com/r/tailscale/tailscale
@Znevna
I was not aware of the TailScale official container [excellent] .. thank you for posting the info …
Useless for me and my tile architecture 
perhaps I can host it on my hex device behind the router.
@anav
Pro Custodibus also have a Docker Container.
https://hub.docker.com/r/procustodibus/agent
and some very good info by Pro Custodibus on containers …
https://www.procustodibus.com/blog/2021/11/wireguard-containers/
I have no idea how this impacts Tik memory … What I like about Pro Custodibus is their amazing and VERY helpful documentation …
I don’t know if you’ve seen this, but I just tested it on my RB5009 and it works great!
https://github.com/Fluent-networks/tailscale-mikrotik/
I advise doing b on point 6. Make sure to edit build.sh and change the platform to the correct one (linux/arm64 for RB5009) and the Tailscale version to the latest.
I installed it as a guide, the image is already running, but the tailscale management website doesn’t show the node. Please share how to do, configure your tailscale container. Thank you!