Port mapper auto discovery used to run immediately after parsing the
--nat flag, giving it a slight performance boost. But this is becoming
inconvenient because we create node.Node for all geth operations
including account management and bare chain interaction. Delay
autodiscovery until the first use instead, which avoids any network
interaction until the node is actually started.

This silences a go vet message about copying p2p.Server in package node.

With the Ethereum Homestead fork is now behind us, we can
assume that everyone runs an EIP-8 capable client.

As we aren't really using the standarized SHA-3

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

The test expected the timeout to fire after a matcher for the response
was added, but the timeout is random and fired sooner sometimes.

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.

This change simplifies the dial scheduling logic because it
no longer needs to track whether the discovery table has been
bootstrapped.

thanks to Felix Lange (fjl) for help with design & impl

The strict matching can get in the way of protocol upgrades.

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.

This reverts commit 5c949d3b.

PR #1621 changed Table locking so the mutex is not held while a
contested node is being pinged. If multiple nodes ping the local node
during this time window, multiple ping packets will be sent to the
contested node. The changes in this commit prevent multiple packets by
tracking whether the node is being replaced.

Might solve #1579

We had the wrong value (12) since forever.

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.

Table.mutex was being held while waiting for a reply packet, which
effectively made many parts of the whole stack block on that packet,
including the net_peerCount RPC call.

Not closing the table used to be fine, but now the table has a database.

Package fdtrack logs statistics about open file descriptors.
This should help identify the source of #1549.