feat: deploy js-libp2p bootstrapper

[SgtPooki]

This issue is tracking the deployment of this bootstrap on our bifrost infra

discussed at https://pl-strflt.notion.site/mob-create-js-libp2p-based-Amino-DHT-bootstrapper-e7793d5a79764df6a08dbc0474935c4d?pvs=4

Tasks

Beta Give feedback

1. Initial alpha deployment feat: bootstrapper ready alpha use/testing in deployed env #135
2. Figure out why memory use is slowly increasing
3. Get a domain name
4. deploy with terraform and add to waterworks monitoring

[SgtPooki]

unassigning this so I can stay focused on Helia and IPFS stuff

[dhuseby]

@wemeetagain was going to try to deploy on chainsafe infra but got stalled.

[dhuseby]

@SgtPooki to deploy a bootstrapper on sgtpooki.com

[SgtPooki]

it could use some robustness and productionization, but it's deployed:

╰─ ✔ ❯ ./vole libp2p identify /ip4/147.75.198.229/tcp/4003/ws/p2p/12D3KooWQgrExcg6dkCdiTER3G3ARe14PZ4cLhinKtRcLHsvnk1Q                                  14.88   17.1G   64%   78%  ─╯
PeerID: "12D3KooWQgrExcg6dkCdiTER3G3ARe14PZ4cLhinKtRcLHsvnk1Q"
Protocol version: "ipfs/0.1.0"
Agent version: "libp2p/1.5.1 UserAgent=v20.13.0"
Listen addresses:
	- "/ip4/147.75.198.229/tcp/4001/p2p/12D3KooWQgrExcg6dkCdiTER3G3ARe14PZ4cLhinKtRcLHsvnk1Q"
	- "/ip4/147.75.198.229/tcp/4003/ws/p2p/12D3KooWQgrExcg6dkCdiTER3G3ARe14PZ4cLhinKtRcLHsvnk1Q"
Protocols:
	- "/ipfs/id/1.0.0"
	- "/ipfs/id/push/1.0.0"
	- "/ipfs/kad/1.0.0"
	- "/libp2p/autonat/1.0.0"
	- "/libp2p/circuit/relay/0.2.0/hop"
	- "/webrtc-signaling/0.0.1"

Fixed by #135

[SgtPooki]

I got some graphs using the metrics dashboards from helia-http-gateway:

[SgtPooki]

graph of last 24 hours (not full because it was only started yesterday afternoon) seems to show steadily increasing memory.

[achingbrain]

Does it need autonat? We can pre-configure it with public announce addresses, right? No need for it to try to detect what these are.

Though I guess we want it to function as a verifier for other autonat peers.

[SgtPooki]

We could just announce the ip/domain for sure, but as a bootstrap node I think it makes sense to help any "startup" things needed by peers. I had autonat disabled at first, but enabled it when docker wasn't working with port publishing, and then just left it when switching to --network host

here's nodejs metrics from the same helia-http-gateway dashboards (-old is the blue increasing line):

[SgtPooki]

FYI @achingbrain it looks like they may be correlated with TCP listener errors?

[2color]

Weirdly, I cannot resolve the peerID from the delegated routing endpoint: https://delegated-ipfs.dev/routing/v1/peers/12D3KooWQgrExcg6dkCdiTER3G3ARe14PZ4cLhinKtRcLHsvnk1Q

[SgtPooki]

Weirdly, I can resolve the peerID from the delegated routing endpoint: delegated-ipfs.dev/routing/v1/peers/12D3KooWQgrExcg6dkCdiTER3G3ARe14PZ4cLhinKtRcLHsvnk1Q

Can or cannot? It's only returning {"Peers":null} for me right now, but some of the DHT provide issues Alex recently fixed (talked about during Helia WG) may be related?

If "can" is correct, why is that weird?

[2color]

I meant that I cannot resolve.