chore(testing): use random seed instead of hardcoding

[Rexicon226]

It's preferred to use the random seed provided by the stdlib (which changes every run, but is deterministic, and can be passed into the test runner again to reproduce identical results), instead of just hardcoding a single number. This ensures the unit doesn't have some hidden dependence on the seed of the prng used.

Had to fix a couple cases just like this, where we would hardcode expected results which depend on the seed to the prng. The change makes it clearer what inputs are used to get said output.

Almost the entire diff is just that, maybe a few little fixes in random places.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
src/accountsdb/db.zig	88.91% <100.00%> (ø)
src/accountsdb/manager.zig	97.09% <100.00%> (ø)
src/accountsdb/snapshot/download.zig	56.46% <100.00%> (ø)
src/accountsdb/swiss_map.zig	97.04% <100.00%> (ø)
src/consensus/fork_choice.zig	98.12% <100.00%> (ø)
src/consensus/latest_validator_votes.zig	95.37% <100.00%> (ø)
src/consensus/replay_tower.zig	93.73% <100.00%> (ø)
src/consensus/vote_listener.zig	92.31% <100.00%> (ø)
src/consensus/vote_tracker.zig	73.85% <100.00%> (ø)
src/core/blockhash_queue.zig	100.00% <100.00%> (ø)
... and 72 more

... and 3 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.