def some_checkpoint(random: SimRandom, view_no: int,
pp_seq_no: int) -> Checkpoint:
return Checkpoint(instId=0,
viewNo=view_no,
seqNoStart=pp_seq_no,
seqNoEnd=pp_seq_no,
digest=base58.b58encode(random.string(32)).decode())
def some_checkpoint(random: SimRandom, view_no: int,
pp_seq_no: int) -> Checkpoint:
return Checkpoint(instId=0,
viewNo=view_no,
seqNoStart=pp_seq_no,
seqNoEnd=pp_seq_no,
digest=random.string(40))
def generate_some(rnd: SimRandom):
values = [rnd.integer(-37, 7342) for _ in range(1000)]
strings = [rnd.string(10) for _ in range(10)]
choice = rnd.choice(*values)
sample = rnd.sample(values, 10)
shuffled = rnd.shuffle(values)
return values, strings, choice, sample, shuffled
def generate_some(rnd: SimRandom):
integers = [rnd.integer(-37, 7342) for _ in range(1000)]
floats = [rnd.float(0.1, 5.5) for _ in range(1000)]
strings = [rnd.string(10) for _ in range(10)]
choice = rnd.choice(*integers)
sample = rnd.sample(integers, 10)
shuffled = rnd.shuffle(integers)
return integers, floats, strings, choice, sample, shuffled
def some_pool(random: SimRandom) -> (SimPool, List):
pool_size = random.integer(4, 8)
pool = SimPool(pool_size, random)
# Create simulated history
# TODO: Move into helper?
faulty = (pool_size - 1) // 3
seq_no_per_cp = 10
max_batches = 50
batches = [BatchID(0, n, random.string(40)) for n in range(1, max_batches)]
checkpoints = [
some_checkpoint(random, 0, n)
for n in range(0, max_batches, seq_no_per_cp)
]
# Preprepares
pp_count = [random.integer(0, len(batches)) for _ in range(pool_size)]
max_pp = sorted(pp_count)[faulty]
# Prepares
p_count = [random.integer(0, min(max_pp, pp)) for pp in pp_count]
max_p = sorted(p_count)[faulty]
# Checkpoints
cp_count = [
1 + random.integer(0, min(max_p, p)) // seq_no_per_cp for p in pp_count
]
max_stable_cp_indx = sorted(cp_count)[faulty] - 1
stable_cp = [
checkpoints[random.integer(0, min(max_stable_cp_indx, cp))].seqNoEnd
for cp in cp_count
]
# Initialize consensus data
for i, node in enumerate(pool.nodes):
node._data.preprepared = batches[:pp_count[i]]
node._data.prepared = batches[:p_count[i]]
node._data.checkpoints = checkpoints[:cp_count[i]]
node._data.stable_checkpoint = stable_cp[i]
committed = []
for i in range(1, max_batches):
prepare_count = sum(1 for node in pool.nodes
if i <= len(node._data.prepared))
has_prepared_cert = prepare_count >= pool_size - faulty
if has_prepared_cert:
committed.append(batches[i - 1])
return pool, committed
def some_pool(random: SimRandom) -> (SimPool, List):
pool_size = random.integer(4, 8)
pool = SimPool(pool_size, random)
view_no = pool._initial_view_no
log_size = pool.nodes[0].config.LOG_SIZE
# Create simulated history
# TODO: Move into helper?
faulty = (pool_size - 1) // 3
seq_no_per_cp = 10
max_batches = 50
batches = [
BatchID(view_no, view_no, n, random.string(40))
for n in range(1, max_batches)
]
checkpoints = [
some_checkpoint(random, view_no, n)
for n in range(0, max_batches, seq_no_per_cp)
]
# Preprepares
pp_count = [random.integer(0, len(batches)) for _ in range(pool_size)]
max_pp = sorted(pp_count)[faulty]
# Prepares
p_count = [random.integer(0, min(max_pp, pp)) for pp in pp_count]
max_p = sorted(p_count)[faulty]
# Checkpoints
cp_count = [
1 + random.integer(0, min(max_p, p)) // seq_no_per_cp for p in pp_count
]
max_stable_cp_indx = sorted(cp_count)[faulty]
stable_cp = [
checkpoints[random.integer(0,
min(max_stable_cp_indx, cp) - 1)].seqNoEnd
for cp in cp_count
]
# Initialize consensus data
for i, node in enumerate(pool.nodes):
high_watermark = stable_cp[i] + log_size
node._data.preprepared = batches[:min(high_watermark, pp_count[i])]
node._data.prepared = batches[:min(high_watermark, p_count[i])]
node._data.checkpoints.clear()
node._data.checkpoints.update(checkpoints[:cp_count[i]])
node._data.stable_checkpoint = stable_cp[i]
# Mock Ordering service to update preprepares for new view
for node in pool.nodes:
def update_shared_data(node, msg: NewViewCheckpointsApplied):
x = [
BatchID(view_no=msg.view_no,
pp_view_no=batch_id.pp_view_no,
pp_seq_no=batch_id.pp_seq_no,
pp_digest=batch_id.pp_digest)
for batch_id in msg.batches
]
node._orderer._data.preprepared = x
node._orderer._subscription.subscribe(
node._orderer._stasher, NewViewCheckpointsApplied,
partial(update_shared_data, node))
committed = []
for i in range(1, max_batches):
prepare_count = sum(1 for node in pool.nodes
if i <= len(node._data.prepared))
has_prepared_cert = prepare_count >= pool_size - faulty
if has_prepared_cert:
batch_id = batches[i - 1]
committed.append(
BatchID(batch_id.view_no + 1, batch_id.pp_view_no,
batch_id.pp_seq_no, batch_id.pp_digest))
return pool, committed
def some_random_preprepare(random: SimRandom, view_no: int,
pp_seq_no: int) -> PrePrepare:
return some_preprepare(view_no, pp_seq_no, random.string(40))