tinc - Aug 2014 - Purging info about a deleted node

Hello all,

this is just a minor nuisance but nevertheless I would not dislike a 
solution for it.

I run a mixed 1.0/1.1 VPN. Some time ago I've removed a node from the 
VPN (wiped the disk, so deleting the private keys). Over the time I've 
removed the corresponding public keys (among with all the infos related 
to the node) from the other nodes. All nodes have been restarted more 
than once since the deletion of the files related to the removed node.

Problem is that as soon as a node rejoins the VPN, it gets the info 
about the removed node from the VPN so that I see the "ghost" node 
listed every time I do a `dump nodes` or `dump graph`.
Shutting down all the nodes at the same time just for forgetting the 
removed node is not an option, so I was wondering if there is (or is 
planned) a way for forcing a purge of such infos.

Also, I would consider the distributed "persisting" infos about the 
deleted node a feature given the mesh nature of tinc. Anyway, the output 
of `tinc info` let me wonder if it really is:

Node:         ghost
Address:      unknown port unknown
Last seen:    never
Status:       indirect
Options:
Protocol:     17.0
Reachability: unreachable
Edges:
Subnets:

As there is no info about the subnets, I doubt the data about this node 
is useful in any way. Also I'm not sure about the accuracy of the 
protocol version, as I'm almost sure it was running a 1.1 release.

Ciao,
Gian Piero.

On Wed, Aug 06, 2014 at 09:24:32PM +0200, Gian Piero Carrubba wrote:
> this is just a minor nuisance but nevertheless I would not dislike a
> solution for it.
> 
> I run a mixed 1.0/1.1 VPN. Some time ago I've removed a node from the
VPN
> (wiped the disk, so deleting the private keys). Over the time I've
removed
> the corresponding public keys (among with all the infos related to the
node)
> from the other nodes. All nodes have been restarted more than once since
the
> deletion of the files related to the removed node.
> 
> Problem is that as soon as a node rejoins the VPN, it gets the info about
> the removed node from the VPN so that I see the "ghost" node
listed every
> time I do a `dump nodes` or `dump graph`.
A solution is already available. You can remove these ghosts by running
this command on one of the nodes that is currently online:

tinc -n <netname> purge

You can achieve the same with tinc 1.0.x by running:

tincd -n <netname> -kWINCH

-- 
Met vriendelijke groet / with kind regards,
     Guus Sliepen <guus at tinc-vpn.org>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL:
<http://www.tinc-vpn.org/pipermail/tinc/attachments/20140807/b307d45f/attachment.sig>