Highlights
Smart Dialing
In our last release, we shipped Smart Dialing. To reiterate, it’s a clever way to reduce the number of spurious dials. Instead of dialing all addresses in parallel (which is what we did before v0.28), we now carefully rank the addresses and dial them one by one.
However, there were two areas where the logic we introduced could lead to suboptimal results:
- There are some networks that block UDP. This means that we won’t be able to dial any QUIC or WebTransport connections at all. This is problematic since our smart dialing logic dials a QUIC address first before dialing a TCP address, which would lead to a regression for these users.
- Similarly, not all ISPs support IPv6 yet. This is problematic as well, since we prefer IPv6 addresses over IPv4.
This is why smart dialing was disabled by default in v0.28.
For this release, we implemented a logic we call Black Hole Detection in this release. We now detect if UDP or IPv6 is completely blocked (blackholed) by observing the outcomes of dial attempts. When a black hole is detected, we skip QUIC / WebTransport and IPv6 addresses, respectively. This significantly reduces the number of unsuccessful dials for users in these environments. To monitor the status of black hole detection, we added two Grafana dashboards to our swarm dashboard, showing the percentage successful / failed dials on UDP and IPv6, and if that leads to our blackhole logic kicking in.
This now allows us to use RFC8305 Happy Eyballs for QUIC: When we dial a peer that has a QUIC IPv6 and IPv4 address, we dial the IPv6 address first, and only start dialing the IPv4 address if we haven’t heard back from the peer within 250ms. Only if we don’t hear back within another 250ms, we start dialing on TCP.
In a future release, we will enable a similar logic for TCP IPv6 and IPv4 addresses, however, this will require us to refactor our code a bit (see #2394 for details).
Preliminary measurements on the IPFS network show that:
- In ~90% of the cases, we end up with a connection on the first address we dial.
- Canceled connection attempts are reduced by more than 60% (compared to the v0.27 release).
Smart Dialing is now enabled by default. We don’t expect it to cause any performance regression, but if you find any problems, please open an issue. It can be disabled via a constructor option to libp2p.New
:
libp2p.SwarmOpts(swarm.WithDialRanker(swarm.NoDelayDialRanker))
Metrics
- It’s now possible to spin up a local Grafana dashboards - with all libp2p dashboards imported and ready to go - by just running
docker-compose up
in thedashboards
directory: #2383. - Resource Manager metrics are now enabled by default: rcmgr: register prometheus metrics with the libp2p registerer by sukunrt · Pull Request #2370 · libp2p/go-libp2p · GitHub.
Changelog
- github.com/libp2p/go-libp2p:
- webtransport: reject listening on a multiaddr with a certhash (#2426) (libp2p/go-libp2p#2426)
- swarm: deprecate libp2p.DialRanker option (#2430) (libp2p/go-libp2p#2430)
- chore: update deps ahead of the v0.29 release (#2427) (libp2p/go-libp2p#2427)
- quic: Update to quic-go v0.36.2 (#2424) (libp2p/go-libp2p#2424)
- autonat: fix typo in WithSchedule option comment (#2425) (libp2p/go-libp2p#2425)
- identify: filter nat64 well-known prefix ipv6 addresses (#2392) (libp2p/go-libp2p#2392)
- update go-multiaddr to v0.10.1, use Unique function from there (#2407) (libp2p/go-libp2p#2407)
- swarm: enable smart dialing by default (#2420) (libp2p/go-libp2p#2420)
- transport integration tests: make TestMoreStreamsThanOurLimits less flaky (#2410) (libp2p/go-libp2p#2410)
- holepunch: skip racy TestDirectDialWorks (#2419) (libp2p/go-libp2p#2419)
- swarm: change relay dial delay to 500ms (#2421) (libp2p/go-libp2p#2421)
- identify: disable racy TestLargeIdentifyMessage with race detector (#2401) (libp2p/go-libp2p#2401)
- swarm: make black hole detection configurable (#2403) (libp2p/go-libp2p#2403)
- net/mock: support ConnectionGater in MockNet (#2297) (libp2p/go-libp2p#2297)
- docs: Add a Github workflow for checking dead links (#2406) (libp2p/go-libp2p#2406)
- rcmgr: enable metrics by default (#2389) (#2409) (libp2p/go-libp2p#2409)
- chore: remove outdated info in README and link to libp2p-implementers slack (#2405) (libp2p/go-libp2p#2405)
- metrics: deduplicate code in examples (#2404) (libp2p/go-libp2p#2404)
- transport tests: remove mplex tests (#2402) (libp2p/go-libp2p#2402)
- swarm: implement Happy Eyeballs ranking (#2365) (libp2p/go-libp2p#2365)
- docs: fix some comments (#2391) (libp2p/go-libp2p#2391)
- metrics: provide separate docker-compose files for OSX and Linux (#2397) (libp2p/go-libp2p#2397)
- chore: Update .github/workflows/stale.yml [skip ci]
- identify: use zero-alloc slice sorting function (#2396) (libp2p/go-libp2p#2396)
- rcmgr: move StatsTraceReporter to rcmgr package (#2388) (libp2p/go-libp2p#2388)
- swarm: implement blackhole detection (#2320) (libp2p/go-libp2p#2320)
- basichost / blankhost: wrap errors (#2331) (libp2p/go-libp2p#2331)
- network: don’t allocate in DedupAddrs (#2395) (libp2p/go-libp2p#2395)
- rcmgr: test snapshot defaults and that we keep consistent defaults (#2315) (libp2p/go-libp2p#2315)
- rcmgr: register prometheus metrics with the libp2p registerer (#2370) (libp2p/go-libp2p#2370)
- metrics: make it possible to spin up Grafana using docker-compose (#2383) (libp2p/go-libp2p#2383)
- identify: set stream deadlines for Identify and Identify Push streams (#2382) (libp2p/go-libp2p#2382)
- fix: in the swarm move Connectedness emit after releasing conns (#2373) (libp2p/go-libp2p#2373)
- metrics: add example for metrics and dashboard (#2232) (libp2p/go-libp2p#2232)
- dashboards: finish metrics effort (#2362) (libp2p/go-libp2p#2362)
- transport tests: many streams and lots of data (#2296) (libp2p/go-libp2p#2296)
- webtransport: close the challenge stream after the Noise handshake (#2305) (libp2p/go-libp2p#2305)
- test: document why InstantTimer is required (#2351) (libp2p/go-libp2p#2351)
- examples: update go-libp2p to v0.28.0 (#2366) (libp2p/go-libp2p#2366)
- rcmgr: fix link to dashboards in README (#2363) (libp2p/go-libp2p#2363)
- chore: Update .github/workflows/stale.yml [skip ci]
- docs: fix some comments errors (#2356) (libp2p/go-libp2p#2356)
- github.com/libp2p/go-yamux/v4 (v4.0.0 → v4.0.1):
- Release v4.0.1 (libp2p/go-yamux#106)
- fix: sendWindowUpdate respects deadlines (#105) (libp2p/go-yamux#105)
- chore: Update .github/workflows/stale.yml [skip ci]
- GitHub - multiformats/go-multiaddr: Composable and future-proof network addresses (v0.9.0 → v0.10.1):
- release v0.10.1 (#206) (multiformats/go-multiaddr#206)
- fix nat64 well-known prefix check (#205) (multiformats/go-multiaddr#205)
- release v0.10.0 (#204) (multiformats/go-multiaddr#204)
- add a Unique function (#203) (multiformats/go-multiaddr#203)
- manet: add function to test if address is NAT64 IPv4 converted IPv6 address (#202) (multiformats/go-multiaddr#202)
- GitHub - multiformats/go-multihash: Multihash implementation in Go (v0.2.2 → v0.2.3):
- chore: release v0.2.3
- perf: outline logic in Decode to allow for stack allocations
Contributors
Contributor | Commits | Lines ± | Files Changed |
---|---|---|---|
Sukun | 14 | +1594/-711 | 48 |
Marco Munizaga | 9 | +1085/-541 | 48 |
Marten Seemann | 14 | +428/-205 | 44 |
Jorropo | 4 | +178/-51 | 8 |
Adrian Sutton | 1 | +190/-17 | 4 |
VM | 2 | +80/-79 | 49 |
Sahib Yar | 1 | +43/-4 | 5 |
Hlib Kanunnikov | 2 | +17/-14 | 5 |
libp2p-mgmt-read-write[bot] | 1 | +26/-0 | 1 |
GitHub | 2 | +6/-19 | 2 |
Prithvi Shahi | 1 | +3/-3 | 1 |
Bryan White | 1 | +2/-2 | 1 |