This fixes an issue where the mumbai testnet node struggle to find
peers. Before this fix in general test peer numbers are typically around
20 in total between eth66, eth67 and eth68. For new peers some can
struggle to find even a single peer after days of operation.
These are the numbers after 12 hours or running on a node which
previously could not find any peers: eth66=13, eth67=76, eth68=91.
The root cause of this issue is the following:
- A significant number of mumbai peers around the boot node return
network ids which are different from those currently available in the
DHT
- The available nodes are all consequently busy and return 'too many
peers' for long periods
These issues case a significant number of discovery timeouts, some of
the queries will never receive a response.
This causes the discovery read loop to enter a channel deadlock - which
means that no responses are processed, nor timeouts fired. This causes
the discovery process in the node to stop. From then on it just
re-requests handshakes from a relatively small number of peers.
This check in fixes this situation with the following changes:
- Remove the deadlock by running the timer in a separate go-routine so
it can run independently of the main request processing.
- Allow the discovery process matcher to match on port if no id match
can be established on initial ping. This allows subsequent node
validation to proceed and if the node proves to be valid via the
remainder of the look-up and handshake process it us used as a valid
peer.
- Completely unsolicited responses, i.e. those which come from a
completely unknown ip:port combination continue to be ignored.
-
* use semaphore instead of a chan struct{}
* move MaxPendingPeers default value to DefaultConfig.P2P
* log Error if Accept fails
* replace quit channel with context
* implemented crash reporting for all goroutine panics that aren't handled explicitly
* implemented crash reporting for all goroutine panics that aren't handled explicitly
* changed node defaults back to originals after testing
* implemented panic handling for all goroutines that don't explicitly handle them, outputting the stack trace to a file in crashreports
* handling panics on all goroutines gracefully
* updated missing call
* error assignment
* implemented suggestions
* path.Join added
* implemented Evgeny's suggestions
* changed path.Join to filepath.Join for cross-platform
* added err check
* updated RecoverStackTrace to LogPanic
* updated closures
* removed call of common.Go to some goroutines
* updated scope capture
* removed testing files
* reverted back to original method, I feel like its less intrusive
* update filename for clarity
* p2p: add low port check in dialer
We already have a check like this for UDP ports, add a similar one in
the dialer. This prevents dials to port zero and it's also an extra
layer of protection against spamming HTTP servers.
* p2p/discover: use errLowPort in v4 code
* p2p: change port check
* p2p: add comment
* p2p/simulations/adapters: ensure assigned port is in all node records
* p2p: new dial scheduler
This change replaces the peer-to-peer dial scheduler with a new and
improved implementation. The new code is better than the previous
implementation in two key aspects:
- The time between discovery of a node and dialing that node is
significantly lower in the new version. The old dialState kept
a buffer of nodes and launched a task to refill it whenever the buffer
became empty. This worked well with the discovery interface we used to
have, but doesn't really work with the new iterator-based discovery
API.
- Selection of static dial candidates (created by Server.AddPeer or
through static-nodes.json) performs much better for large amounts of
static peers. Connections to static nodes are now limited like dynanic
dials and can no longer overstep MaxPeers or the dial ratio.
* p2p/simulations/adapters: adapt to new NodeDialer interface
* p2p: re-add check for self in checkDial
* p2p: remove peersetCh
* p2p: allow static dials when discovery is disabled
* p2p: add test for dialScheduler.removeStatic
* p2p: remove blank line
* p2p: fix documentation of maxDialPeers
* p2p: change "ok" to "added" in static node log
* p2p: improve dialTask docs
Also increase log level for "Can't resolve node"
* p2p: ensure dial resolver is truly nil without discovery
* p2p: add "looking for peers" log message
* p2p: clean up Server.run comments
* p2p: fix maxDialedConns for maxpeers < dialRatio
Always allocate at least one dial slot unless dialing is disabled using
NoDial or MaxPeers == 0. Most importantly, this fixes MaxPeers == 1 to
dedicate the sole slot to dialing instead of listening.
* p2p: fix RemovePeer to disconnect the peer again
Also make RemovePeer synchronous and add a test.
* p2p: remove "Connection set up" log message
* p2p: clean up connection logging
We previously logged outgoing connection failures up to three times.
- in SetupConn() as "Setting up connection failed addr=..."
- in setupConn() with an error-specific message and "id=... addr=..."
- in dial() as "Dial error task=..."
This commit ensures a single log message is emitted per failure and adds
"id=... addr=... conn=..." everywhere (id= omitted when the ID isn't
known yet).
Also avoid printing a log message when a static dial fails but can't be
resolved because discv4 is disabled. The light client hit this case all
the time, increasing the message count to four lines per failed
connection.
* p2p: document that RemovePeer blocks
This adds all dashboard changes from the last couple months.
We're about to remove the dashboard, but decided that we should
get all the recent work in first in case anyone wants to pick up this
project later on.
* cmd, dashboard, eth, p2p: send peer info to the dashboard
* dashboard: update npm packages, improve UI, rebase
* dashboard, p2p: remove println, change doc
* cmd, dashboard, eth, p2p: cleanup after review
* dashboard: send current block to the dashboard client
The dialer limits itself to one attempt every 30s. Apply the same limit
in Server and reject peers which try to connect too eagerly. The check
against the limit happens right after accepting the connection.
Further changes in this commit ensure we pass the Server logger
down to Peer instances, discovery and dialState. Unit test logging now
works in all Server tests.
This change extends the peer metrics collection:
- traces the life-cycle of the peers
- meters the peer traffic separately for every peer
- creates event feed for the peer events
- emits the peer events
This PR adds enode.LocalNode and integrates it into the p2p
subsystem. This new object is the keeper of the local node
record. For now, a new version of the record is produced every
time the client restarts. We'll make it smarter to avoid that in
the future.
There are a couple of other changes in this commit: discovery now
waits for all of its goroutines at shutdown and the p2p server
now closes the node database after discovery has shut down. This
fixes a leveldb crash in tests. p2p server startup is faster
because it doesn't need to wait for the external IP query
anymore.
Package p2p/enode provides a generalized representation of p2p nodes
which can contain arbitrary information in key/value pairs. It is also
the new home for the node database. The "v4" identity scheme is also
moved here from p2p/enr to remove the dependency on Ethereum crypto from
that package.
Record signature handling is changed significantly. The identity scheme
registry is removed and acceptable schemes must be passed to any method
that needs identity. This means records must now be validated explicitly
after decoding.
The enode API is designed to make signature handling easy and safe: most
APIs around the codebase work with enode.Node, which is a wrapper around
a valid record. Going from enr.Record to enode.Node requires a valid
signature.
* p2p/discover: port to p2p/enode
This ports the discovery code to the new node representation in
p2p/enode. The wire protocol is unchanged, this can be considered a
refactoring change. The Kademlia table can now deal with nodes using an
arbitrary identity scheme. This requires a few incompatible API changes:
- Table.Lookup is not available anymore. It used to take a public key
as argument because v4 protocol requires one. Its replacement is
LookupRandom.
- Table.Resolve takes *enode.Node instead of NodeID. This is also for
v4 protocol compatibility because nodes cannot be looked up by ID
alone.
- Types Node and NodeID are gone. Further commits in the series will be
fixes all over the the codebase to deal with those removals.
* p2p: port to p2p/enode and discovery changes
This adapts package p2p to the changes in p2p/discover. All uses of
discover.Node and discover.NodeID are replaced by their equivalents from
p2p/enode.
New API is added to retrieve the enode.Node instance of a peer. The
behavior of Server.Self with discovery disabled is improved. It now
tries much harder to report a working IP address, falling back to
127.0.0.1 if no suitable address can be determined through other means.
These changes were needed for tests of other packages later in the
series.
* p2p/simulations, p2p/testing: port to p2p/enode
No surprises here, mostly replacements of discover.Node, discover.NodeID
with their new equivalents. The 'interesting' API changes are:
- testing.ProtocolSession tracks complete nodes, not just their IDs.
- adapters.NodeConfig has a new method to create a complete node.
These changes were needed to make swarm tests work.
Note that the NodeID change makes the code incompatible with old
simulation snapshots.
* whisper/whisperv5, whisper/whisperv6: port to p2p/enode
This port was easy because whisper uses []byte for node IDs and
URL strings in the API.
* eth: port to p2p/enode
Again, easy to port because eth uses strings for node IDs and doesn't
care about node information in any way.
* les: port to p2p/enode
Apart from replacing discover.NodeID with enode.ID, most changes are in
the server pool code. It now deals with complete nodes instead
of (Pubkey, IP, Port) triples. The database format is unchanged for now,
but we should probably change it to use the node database later.
* node: port to p2p/enode
This change simply replaces discover.Node and discover.NodeID with their
new equivalents.
* swarm/network: port to p2p/enode
Swarm has its own node address representation, BzzAddr, containing both
an overlay address (the hash of a secp256k1 public key) and an underlay
address (enode:// URL).
There are no changes to the BzzAddr format in this commit, but certain
operations such as creating a BzzAddr from a node ID are now impossible
because node IDs aren't public keys anymore.
Most swarm-related changes in the series remove uses of
NewAddrFromNodeID, replacing it with NewAddr which takes a complete node
as argument. ToOverlayAddr is removed because we can just use the node
ID directly.
This change removes a peer information from dialing history
when peer is removed from static list. It allows to force a
server to re-dial concrete peer if it is needed.
In our case we are running geth node on mobile devices, and
it is common for a network connection to flap on mobile.
Almost every time it flaps or network connection is changed
from cellular to wifi peers are disconnected with read
timeout. And usually it takes 30 seconds (default expiration
timeout) to recover connection with static peers after
connectivity is restored.
This change allows us to reconnect with peers almost
immediately and it seems harmless enough.
p2p/simulations: introduce dialBan
- Refactor simulations/network connection getters to support
avoiding simultaneous dials between two peers If two peers dial
simultaneously, the connection will be dropped to help avoid
that, we essentially lock the connection object with a
timestamp which serves as a ban on dialing for a period of time
(dialBanTimeout).
- The connection getter InitConn can be wrapped and passed to the
nodes via adapters.NodeConfig#Reachable field and then used by
the respective services when they initiate connections. This
massively stablise the emerging connectivity when running with
hundreds of nodes bootstrapping a network.
p2p: add Inbound public method to p2p.Peer
p2p/simulations: Add server id to logs to support debugging
in-memory network simulations when multiple peers are logging.
p2p: SetupConn now returns error. The dialer checks the error and
only calls resolve if the actual TCP dial fails.
This commit introduces a network simulation framework which
can be used to run simulated networks of devp2p nodes. The
intention is to use this for testing protocols, performing
benchmarks and visualising emergent network behaviour.
The p2p packages can now be configured to restrict all communication to
a certain subset of IP networks. This feature is meant to be used for
private networks.
This change makes it possible to add peers without providing their IP
address. The endpoint of the target node is resolved using the discovery
protocol.
