This change clears up confusion around the two ways in which nodes
can be added to the table.
When a neighbors packet is received as a reply to findnode, the nodes
contained in the reply are added as 'seen' entries if sufficient space
is available.
When a ping is received and the endpoint verification has taken place,
the remote node is added as a 'verified' entry or moved to the front of
the bucket if present. This also updates the node's IP address and port
if they have changed.
This change resolves multiple issues around handling of endpoint proofs.
The proof is now done separately for each IP and completing the proof
requires a matching ping hash.
Also remove waitping because it's equivalent to sleep. waitping was
slightly more efficient, but that may cause issues with findnode if
packets are reordered and the remote end sees findnode before pong.
Logging of received packets was hitherto done after handling the packet,
which meant that sent replies were logged before the packet that
generated them. This change splits up packet handling into 'preverify'
and 'handle'. The error from 'preverify' is logged, but 'handle' happens
after the message is logged. This fixes the order. Packet logs now
contain the node ID.
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.
* p2p/discover: move bond logic from table to transport
This commit moves node endpoint verification (bonding) from the table to
the UDP transport implementation. Previously, adding a node to the table
entailed pinging the node if needed. With this change, the ping-back
logic is embedded in the packet handler at a lower level.
It is easy to verify that the basic protocol is unchanged: we still
require a valid pong reply from the node before findnode is accepted.
The node database tracked the time of last ping sent to the node and
time of last valid pong received from the node. Node endpoints are
considered verified when a valid pong is received and the time of last
pong was called 'bond time'. The time of last ping sent was unused. In
this commit, the last ping database entry is repurposed to mean last
ping _received_. This entry is now used to track whether the node needs
to be pinged back.
The other big change is how nodes are added to the table. We used to add
nodes in Table.bond, which ran when a remote node pinged us or when we
encountered the node in a neighbors reply. The transport now adds to the
table directly after the endpoint is verified through ping. To ensure
that the Table can't be filled just by pinging the node repeatedly, we
retain the isInitDone check. During init, only nodes from neighbors
replies are added.
* p2p/discover: reduce findnode failure counter on success
* p2p/discover: remove unused parameter of loadSeedNodes
* p2p/discover: improve ping-back check and comments
* p2p/discover: add neighbors reply nodes always, not just during init
I forgot to change the check in udp.go when I changed Table.bond to be
based on lastPong instead of node presence in db. Rename lastPong to
bondTime and add hasBond so it's clearer what this DB key is used for
now.
* p2p: add DialRatio for configuration of inbound vs. dialed connections
* p2p: add connection flags to PeerInfo
* p2p/netutil: add SameNet, DistinctNetSet
* p2p/discover: improve revalidation and seeding
This changes node revalidation to be periodic instead of on-demand. This
should prevent issues where dead nodes get stuck in closer buckets
because no other node will ever come along to replace them.
Every 5 seconds (on average), the last node in a random bucket is
checked and moved to the front of the bucket if it is still responding.
If revalidation fails, the last node is replaced by an entry of the
'replacement list' containing recently-seen nodes.
Most close buckets are removed because it's very unlikely we'll ever
encounter a node that would fall into any of those buckets.
Table seeding is also improved: we now require a few minutes of table
membership before considering a node as a potential seed node. This
should make it less likely to store short-lived nodes as potential
seeds.
* p2p/discover: fix nits in UDP transport
We would skip sending neighbors replies if there were fewer than
maxNeighbors results and CheckRelayIP returned an error for the last
one. While here, also resolve a TODO about pong reply tokens.
This commit affects p2p/discv5 "topic discovery" by running it on
the same UDP port where the old discovery works. This is realized
by giving an "unhandled" packet channel to the old v4 discovery
packet handler where all invalid packets are sent. These packets
are then processed by v5. v5 packets are always invalid when
interpreted by v4 and vice versa. This is ensured by adding one
to the first byte of the packet hash in v5 packets.
DiscoveryV5Bootnodes is also changed to point to new bootnodes
that are implementing the changed packet format with modified
hash. Existing and new v5 bootnodes are both running on different
ports ATM.
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.
The discovery DHT contains a number of hosts with LAN and loopback IPs.
These get relayed because some implementations do not perform any checks
on the IP.
go-ethereum already prevented relay in most cases because it verifies
that the host actually exists before adding it to the local table. But
this verification causes other issues. We have received several reports
where people's VPSs got shut down by hosting providers because sending
packets to random LAN hosts is indistinguishable from a slow port scan.
The new check prevents sending random packets to LAN by discarding LAN
IPs sent by Internet hosts (and loopback IPs from LAN and Internet
hosts). The new check also blacklists almost all currently registered
special-purpose networks assigned by IANA to avoid inciting random
responses from services in the LAN.
As another precaution against abuse of the DHT, ports below 1024 are now
considered invalid.
On Windows, UDPConn.ReadFrom returns an error for packets larger
than the receive buffer. The error is not marked temporary, causing
our loop to exit when the first oversized packet arrived. The fix
is to treat this particular error as temporary.
Fixes: #1579, #2087
Updates: #2082
nodeDB.querySeeds was not safe for concurrent use but could be called
concurrenty on multiple goroutines in the following case:
- the table was empty
- a timed refresh started
- a lookup was started and initiated refresh
These conditions are unlikely to coincide during normal use, but are
much more likely to occur all at once when the user's machine just woke
from sleep. The root cause of the issue is that querySeeds reused the
same leveldb iterator until it was exhausted.
This commit moves the refresh scheduling logic into its own goroutine
(so only one refresh is ever active) and changes querySeeds to not use
a persistent iterator. The seed node selection is now more random and
ignores nodes that have not been contacted in the last 5 days.
If the timeout fired (even just nanoseconds) before the deadline of the
next pending reply, the timer was not rescheduled. The timer would've
been rescheduled anyway once the next packet was sent, but there were
cases where no next packet could ever be sent due to the locking issue
fixed in the previous commit.
As timing-related bugs go, this issue had been present for a long time
and I could never reproduce it. The test added in this commit did
reproduce the issue on about one out of 15 runs.
We don't have a UDP which specifies any messages that will be 4KB. Aside from being implemented for months and a necessity for encryption and piggy-backing packets, 1280bytes is ideal, and, means this TODO can be completed!
Why 1280 bytes?
* It's less than the default MTU for most WAN/LAN networks. That means fewer fragmented datagrams (esp on well-connected networks).
* Fragmented datagrams and dropped packets suck and add latency while OS waits for a dropped fragment to never arrive (blocking readLoop())
* Most of our packets are < 1280 bytes.
* 1280 bytes is minimum datagram size and MTU for IPv6 -- on IPv6, a datagram < 1280bytes will *never* be fragmented.
UDP datagrams are dropped. A lot! And fragmented datagrams are worse. If a datagram has a 30% chance of being dropped, then a fragmented datagram has a 60% chance of being dropped. More importantly, we have signed packets and can't do anything with a packet unless we receive the entire datagram because the signature can't be verified. The same is true when we have encrypted packets.
So the solution here to picking an ideal buffer size for receiving datagrams is a number under 1400bytes. And the lower-bound value for IPv6 of 1280 bytes make's it a non-decision. On IPv4 most ISPs and 3g/4g/let networks have an MTU just over 1400 -- and *never* over 1500. Never -- that means packets over 1500 (in reality: ~1450) bytes are fragmented. And probably dropped a lot.
Just to prove the point, here are pings sending non-fragmented packets over wifi/ISP, and a second set of pings via cell-phone tethering. It's important to note that, if *any* router between my system and the EC2 node has a lower MTU, the message would not go through:
On wifi w/normal ISP:
localhost:Debug $ ping -D -s 1450 52.6.250.242
PING 52.6.250.242 (52.6.250.242): 1450 data bytes
1458 bytes from 52.6.250.242: icmp_seq=0 ttl=42 time=104.831 ms
1458 bytes from 52.6.250.242: icmp_seq=1 ttl=42 time=119.004 ms
^C
--- 52.6.250.242 ping statistics ---
2 packets transmitted, 2 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 104.831/111.918/119.004/7.087 ms
localhost:Debug $ ping -D -s 1480 52.6.250.242
PING 52.6.250.242 (52.6.250.242): 1480 data bytes
ping: sendto: Message too long
ping: sendto: Message too long
Request timeout for icmp_seq 0
ping: sendto: Message too long
Request timeout for icmp_seq 1
Tethering to O2:
localhost:Debug $ ping -D -s 1480 52.6.250.242
PING 52.6.250.242 (52.6.250.242): 1480 data bytes
ping: sendto: Message too long
ping: sendto: Message too long
Request timeout for icmp_seq 0
^C
--- 52.6.250.242 ping statistics ---
2 packets transmitted, 0 packets received, 100.0% packet loss
localhost:Debug $ ping -D -s 1450 52.6.250.242
PING 52.6.250.242 (52.6.250.242): 1450 data bytes
1458 bytes from 52.6.250.242: icmp_seq=0 ttl=42 time=107.844 ms
1458 bytes from 52.6.250.242: icmp_seq=1 ttl=42 time=105.127 ms
1458 bytes from 52.6.250.242: icmp_seq=2 ttl=42 time=120.483 ms
1458 bytes from 52.6.250.242: icmp_seq=3 ttl=42 time=102.136 ms
The previous metric was pubkey1^pubkey2, as specified in the Kademlia
paper. We missed that EC public keys are not uniformly distributed.
Using the hash of the public keys addresses that. It also makes it
a bit harder to generate node IDs that are close to a particular node.
This commit changes the discovery protocol to use the new "v4" endpoint
format, which allows for separate UDP and TCP ports and makes it
possible to discover the UDP address after NAT.
