test: address self-announcement

[0xB10C]

Test that a node sends a self-announcement with its external IP to in- and outbound peers after connection open and again sometime later.

Since the code for the test is mostly the same for addr and addrv2 messages, I opted to add a new test file instead of having duplicate code in p2p_addr_relay.py and p2p_addrv2_relay.py.

[DrahtBot]

The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Code Coverage & Benchmarks

For details see: https://corecheck.dev/bitcoin/bitcoin/pulls/34039.

Reviews

See the guideline for information on the review process.

Type	Reviewers
ACK	Bicaru20

If your review is incorrectly listed, please copy-paste <!--meta-tag:bot-skip--> into the comment that the bot should ignore.

[0xB10C]

inspired by

bitcoin/test/functional/p2p_addr_relay.py

Lines 130 to 133 in cca113f

	# invoke m_next_addr_send timer:
	# `addr` messages are sent on an exponential distribution with mean interval of 30s.
	# Setting the mocktime 600s forward gives a probability of (1 - e^-(600/30)) that
	# the event will occur (i.e. this fails once in ~500 million repeats).

(using the same exponent of e^-20 here)

[mzumsande]

nothing for this PR, but since these things are quite common (recently #33121), I wonder if a mode -test=cutoffexponentialtimers or something similar would make sense which would cut off the long tail of all exponential timers (rand_exp_duration) at some percentile, so we don't need to deal with these probabilistic failures in functional tests.

[Bicaru20]

Since increasing the 50 in the loop will not contribute anything to the test, maybe the >> 8 and % 256 could be erased.

Suggested change

	first_octet = 1 + (i >> 8)
	second_octet = i % 256
	first_octet = 1 + i
	second_octet = i

[Bicaru20]

Code review ACK fb9c393. I have run the test and everything checks out.

I left a small comment in the code, also fine with merging as-is