kubo/peering/peering_test.go
Marten Seemann 514411bedb
feat: opt-in Swarm.ResourceMgr (go-libp2p v0.18) (#8680)
* update go-libp2p to v0.18.0

* initialize the resource manager

* add resource manager stats/limit commands

* load limit file when building resource manager

* log absent limit file

* write rcmgr to file when IPFS_DEBUG_RCMGR is set

* fix: mark swarm limit|stats as experimental

* feat(cfg): opt-in Swarm.ResourceMgr

This ensures we can safely test the resource manager without impacting
default behavior.

- Resource manager is disabled by default
    - Default for Swarm.ResourceMgr.Enabled is false for now
- Swarm.ResourceMgr.Limits allows user to tweak limits per specific
  scope in a way that is persisted across restarts
- 'ipfs swarm limit system' outputs human-readable json
- 'ipfs swarm limit system new-limits.json' sets new runtime limits
  (but does not change Swarm.ResourceMgr.Limits in the config)

Conventions to make libp2p devs life easier:
- 'IPFS_RCMGR=1 ipfs daemon' overrides the config and enables resource manager
- 'limit.json' overrides implicit defaults from libp2p (if present)

* docs(config): small tweaks

* fix: skip libp2p.ResourceManager if disabled

This ensures 'ipfs swarm limit|stats' work only when enabled.

* fix: use NullResourceManager when disabled

This reverts commit b19f7c9eca.
after clarification feedback from
https://github.com/ipfs/go-ipfs/pull/8680#discussion_r841680182

* style: rename IPFS_RCMGR to LIBP2P_RCMGR

preexisting libp2p toggles use LIBP2P_ prefix

* test: Swarm.ResourceMgr

* fix: location of opt-in limit.json and rcmgr.json.gz

Places these files inside of IPFS_PATH

* Update docs/config.md

* feat: expose rcmgr metrics when enabled (#8785)

* add metrics for the resource manager
* export protocol and service name in Prometheus metrics
* fix: expose rcmgr metrics only when enabled

Co-authored-by: Marcin Rataj <lidel@lidel.org>

* refactor: rcmgr_metrics.go

* refactor: rcmgr_defaults.go

This file defines implicit limit defaults used when Swarm.ResourceMgr.Enabled

We keep vendored copy to ensure go-ipfs is not impacted when go-libp2p
decides to change defaults in any of the future releases.

* refactor: adjustedDefaultLimits

Cleans up the way we initialize defaults and adds a fix for case
when connection manager runs with high limits.

It also hides `Swarm.ResourceMgr.Limits` until we have a better
understanding what syntax makes sense.

* chore: cleanup after a review

* fix: restore go-ipld-prime v0.14.2

* fix: restore go-ds-flatfs v0.5.1

Co-authored-by: Lucas Molas <schomatis@gmail.com>
Co-authored-by: Marcin Rataj <lidel@lidel.org>
2022-04-07 21:06:35 -04:00

173 lines
5.5 KiB
Go

package peering
import (
"context"
"testing"
"time"
"github.com/libp2p/go-libp2p"
"github.com/libp2p/go-libp2p-core/host"
"github.com/libp2p/go-libp2p-core/network"
"github.com/libp2p/go-libp2p-core/peer"
"github.com/libp2p/go-libp2p/p2p/net/connmgr"
"github.com/stretchr/testify/require"
)
func newNode(t *testing.T) host.Host {
cm, err := connmgr.NewConnManager(1, 100, connmgr.WithGracePeriod(0))
require.NoError(t, err)
h, err := libp2p.New(
libp2p.ListenAddrStrings("/ip4/127.0.0.1/tcp/0"),
// We'd like to set the connection manager low water to 0, but
// that would disable the connection manager.
libp2p.ConnectionManager(cm),
)
require.NoError(t, err)
return h
}
func TestPeeringService(t *testing.T) {
ctx, cancel := context.WithCancel(context.Background())
defer cancel()
h1 := newNode(t)
ps1 := NewPeeringService(h1)
h2 := newNode(t)
h3 := newNode(t)
h4 := newNode(t)
// peer 1 -> 2
ps1.AddPeer(peer.AddrInfo{ID: h2.ID(), Addrs: h2.Addrs()})
require.Contains(t, ps1.ListPeers(), peer.AddrInfo{ID: h2.ID(), Addrs: h2.Addrs()})
// We haven't started so we shouldn't have any peers.
require.Never(t, func() bool {
return len(h1.Network().Peers()) > 0
}, 100*time.Millisecond, 1*time.Second, "expected host 1 to have no peers")
// Use p4 to take up the one slot we have in the connection manager.
for _, h := range []host.Host{h1, h2} {
require.NoError(t, h.Connect(ctx, peer.AddrInfo{ID: h4.ID(), Addrs: h4.Addrs()}))
h.ConnManager().TagPeer(h4.ID(), "sticky-peer", 1000)
}
// Now start.
require.NoError(t, ps1.Start())
// starting twice is fine.
require.NoError(t, ps1.Start())
// We should eventually connect.
t.Logf("waiting for h1 to connect to h2")
require.Eventually(t, func() bool {
return h1.Network().Connectedness(h2.ID()) == network.Connected
}, 30*time.Second, 10*time.Millisecond)
// Now explicitly connect to h3.
t.Logf("waiting for h1's connection to h3 to work")
require.NoError(t, h1.Connect(ctx, peer.AddrInfo{ID: h3.ID(), Addrs: h3.Addrs()}))
require.Eventually(t, func() bool {
return h1.Network().Connectedness(h3.ID()) == network.Connected
}, 30*time.Second, 100*time.Millisecond)
require.Len(t, h1.Network().Peers(), 3)
// force a disconnect
h1.ConnManager().TrimOpenConns(ctx)
// Should disconnect from h3.
t.Logf("waiting for h1's connection to h3 to disconnect")
require.Eventually(t, func() bool {
return h1.Network().Connectedness(h3.ID()) != network.Connected
}, 5*time.Second, 10*time.Millisecond)
// Should remain connected to p2
require.Never(t, func() bool {
return h1.Network().Connectedness(h2.ID()) != network.Connected
}, 5*time.Second, 1*time.Second)
// Now force h2 to disconnect (we have an asymmetric peering).
conns := h2.Network().ConnsToPeer(h1.ID())
require.NotEmpty(t, conns)
h2.ConnManager().TrimOpenConns(ctx)
// All conns to peer should eventually close.
t.Logf("waiting for all connections to close")
for _, c := range conns {
require.Eventually(t, func() bool {
s, err := c.NewStream(context.Background())
if s != nil {
_ = s.Reset()
}
return err != nil
}, 5*time.Second, 10*time.Millisecond)
}
// Should eventually re-connect.
require.Eventually(t, func() bool {
return h1.Network().Connectedness(h2.ID()) == network.Connected
}, 30*time.Second, 1*time.Second)
// Unprotect 2 from 1.
ps1.RemovePeer(h2.ID())
require.NotContains(t, ps1.ListPeers(), peer.AddrInfo{ID: h2.ID(), Addrs: h2.Addrs()})
// Trim connections.
h1.ConnManager().TrimOpenConns(ctx)
// Should disconnect
t.Logf("waiting for h1 to disconnect from h2")
require.Eventually(t, func() bool {
return h1.Network().Connectedness(h2.ID()) != network.Connected
}, 5*time.Second, 10*time.Millisecond)
// Should never reconnect.
t.Logf("ensuring h1 is not connected to h2 again")
require.Never(t, func() bool {
return h1.Network().Connectedness(h2.ID()) == network.Connected
}, 20*time.Second, 1*time.Second)
// Until added back
ps1.AddPeer(peer.AddrInfo{ID: h2.ID(), Addrs: h2.Addrs()})
require.Contains(t, ps1.ListPeers(), peer.AddrInfo{ID: h2.ID(), Addrs: h2.Addrs()})
ps1.AddPeer(peer.AddrInfo{ID: h3.ID(), Addrs: h3.Addrs()})
require.Contains(t, ps1.ListPeers(), peer.AddrInfo{ID: h3.ID(), Addrs: h3.Addrs()})
t.Logf("wait for h1 to connect to h2 and h3 again")
require.Eventually(t, func() bool {
return h1.Network().Connectedness(h2.ID()) == network.Connected
}, 30*time.Second, 1*time.Second)
require.Eventually(t, func() bool {
return h1.Network().Connectedness(h3.ID()) == network.Connected
}, 30*time.Second, 1*time.Second)
// Should be able to repeatedly stop.
require.NoError(t, ps1.Stop())
require.NoError(t, ps1.Stop())
// Adding and removing should work after stopping.
ps1.AddPeer(peer.AddrInfo{ID: h4.ID(), Addrs: h4.Addrs()})
require.Contains(t, ps1.ListPeers(), peer.AddrInfo{ID: h4.ID(), Addrs: h4.Addrs()})
ps1.RemovePeer(h2.ID())
require.NotContains(t, ps1.ListPeers(), peer.AddrInfo{ID: h2.ID(), Addrs: h2.Addrs()})
}
func TestNextBackoff(t *testing.T) {
minMaxBackoff := (100 - maxBackoffJitter) / 100 * maxBackoff
for x := 0; x < 1000; x++ {
ph := peerHandler{nextDelay: time.Second}
for min, max := time.Second*3/2, time.Second*5/2; min < minMaxBackoff; min, max = min*3/2, max*5/2 {
b := ph.nextBackoff()
if b > max || b < min {
t.Errorf("expected backoff %s to be between %s and %s", b, min, max)
}
}
for i := 0; i < 100; i++ {
b := ph.nextBackoff()
if b < minMaxBackoff || b > maxBackoff {
t.Fatal("failed to stay within max bounds")
}
}
}
}