That and the some hosts held multiple nodes… I think someone suggested they had nine running and I’m guilty of dropping five at a time; so, perhaps the chance of data loss was quite high at times where large chunks were lost. We were all being rough with it. With more nodes the theoretical risk drops off very quickly, so might be worth seeing a minimum of x50 nodes next time to check there is no real issue.
1 Like
Really want to thank you @bluebird for all the work you’re doing in this area; it’s incredibly important to get these community nets going and I applaud all your continuing effort and work here
7 Likes
just a small point if it wasn’t already mentioned. This config indicates three tcp acceptor endpoints in a single node. Think what you would have been after would have been more like:
{
"hard_coded_contacts": [
{
"tcp_acceptors": [
"91.121.173.204:5483",
],
"tcp_mapper_servers": []
},
{
"tcp_acceptors": [
"88.19.209.226:5483",
],
"tcp_mapper_servers": []
},
{
"tcp_acceptors": [
"207.172.241.242:5483",
],
"tcp_mapper_servers": []
}
],
"tcp_acceptor_port": 5483,
"service_discovery_port": null,
"bootstrap_cache_name": null,
"tcp_mapper_servers": []
}
as for this, there is no set requirements as you basically just need a single endpoint in your config file to bootstrap to the network. Having more contacts in the config file essentially spreads the load amongst bootstrap nodes with incoming peers. If a single node is used by 100 people to try and bootstrap, that node is doing a lot more work in this case.
When a vault/client is run, it picks a random hard-coded contact from it’s config file to bootstrap via. This when bootstrap cache is enabled will further spread the load as its not just endpoints from the config file which will be used to bootstrap to the network.
5 Likes
Hey so has anyone started a community-net yet? Just run Cline command of --first with safe.vault to start one right? Has anyone started one yet?
1 Like
I have started a new community network, called “community1”, but it is just sitting there listening. 
Please copy this into your vaults’ and launchers’ configs, and restart. No command-line options necessary. I’m guessing that that is all that’s needed (or not, lol).
Vaults behind NAT should be fine with the latest hole-punching, but anyone who has a vault in the cloud, connecting directly to community1, feel free to send me your IP and I’ll add it:
{
"hard_coded_contacts": [
{
"tcp_acceptors": [
"91.121.173.204:5483"
],
"tcp_mapper_servers": []
}
],
"tcp_acceptor_port": 5483,
"service_discovery_port": null,
"bootstrap_cache_name": null,
"tcp_mapper_servers": [],
"network_name": "community1"
}
I still have a Test 3 vault running on the same Ip and listening on the same port. I don’t know if that will cause conflict. I hope that each vault will simply reject messages intended for the other’s network, until I take the first one down.
Until it builds up a routing table, launchers won’t be able to connect.
I ran it with the --first option, still just listening.
WHOA, it is connecting (to one other vault)! Taking the Test 3 vault down and adding the --first flag did it.
Just to reiterate: don’t use the --first flag, because there can only be one seed. Just replace your config and that’s it.
Two new entries in the routing table!
Don’t forget to use the current distributables. The earlier versions are rejected. the current ones are here:
SAFE Vault binaries
SAFE Launcher binaries
SAFE Demonstration Application binaries
There are seven entries now. Could someone run a launcher and see if they connect?
4 Likes
let me download a launcher on the other pc - one moment =) (but get requests received: 19)
ps: 56
2 Likes
“Logged in and connected to the SAFE Network”
you are my hero =)
2 Likes
Managed to create a account and upload a test site.
Test mp3 http://test.xorkrik.safenet/02%20-%20The%20USS%20Make%20Sh-t%20Up.mp3
Also running a few vaults.
3 Likes
It was very fast to register and create a site.
3 Likes
My routing table is at 12 right now.
2 Likes
1 Like
@bluebird do update the OP with the current detail, ip address etc, in case people can’t see your newer post.
2 Likes
I’m wondering there is a odd feature that starting a vault suggests it might take a while and putting another one up in parallel, sees the second one is much quicker… then taking the first one down before it makes progress, is a faster route to getting one vault going. I don’t know why that would be but works for me.
1 Like
Just to note my error… we need to also update the safe_launcher.crust.config otherwise a vault runs ok but the launcher then looks back to the network it knew which in this case will be testnet3 - thanks for the headsup @riddim I wonder there is an option to delete the safe_launcher.crust.config and it’ll copy itself from the vault detail but haven’t tried that.
where is the safe_launcher.crust.config hidden on my mac?
vault connected nicely,:
INFO 22:27:04.666993366 [routing::core core.rs:414] -------------------------------------------------------
INFO 22:27:04.666998945 [routing::core core.rs:416] | Node(356a…) PeerId(fc46…) - Routing Table size: 15 |
INFO 22:27:04.667003074 [routing::core core.rs:417] -------------------------------------------------------
rup
In the safe_launcher folder… or if vanilla it might not be there at all and gets copied from the vault config.
umm… my vault is on a unbuntu box, i’m assuming i just need to copy the community1 contents into is when i find it… 
so on the mac there is a single app which i have looked through and can’t see the safe_launcher.crust.config.
rup
Very fast launcher and demo app, with only 14 nodes in the table. The website I put up an hour ago is still there.
2 Likes
got 15 in table… i’ll leave running from now…
would like to suss changing the config on my mac’s so i can try the launcher on community1…
rup
dropping a config file in the same folder as the app seems to do the trick…
now connected via the launcher to com1… hurrah!
rup
3 Likes