compute <-> sk protocol v3 (#10647)

## Problem

As part of https://github.com/neondatabase/neon/issues/8614 we need to
pass membership configurations between compute and safekeepers.

## Summary of changes

Add version 3 of the protocol carrying membership configurations.
Greeting message in both sides gets full conf, and other messages
generation number only. Use protocol bump to include other accumulated
changes:
- stop packing whole structs on the wire as is;
- make the tag u8 instead of u64;
- send all ints in network order;
- drop proposer_uuid, we can pass it in START_WAL_PUSH and it wasn't
much useful anyway.
Per message changes, apart from mconf:
- ProposerGreeting: tenant / timeline id is sent now as hex cstring.
Remove proto version, it is passed outside in START_WAL_PUSH. Remove
postgres timeline, it is unused. Reorder fields a bit.
- AcceptorGreeting: reorder fields
- VoteResponse: timeline_start_lsn is removed. It can be taken from
first member of term history, and later we won't need it at all when all
timelines will be explicitly created. Vote itself is u8 instead of u64.
- ProposerElected: timeline_start_lsn is removed for the same reasons.
- AppendRequest: epoch_start_lsn removed, it is known from term history
in ProposerElected.

Both compute and sk are able to talk v2 and v3 to make rollbacks (in
case we need them) easier; neon.safekeeper_proto_version GUC sets the
client version. v2 code can be dropped later.

So far empty conf is passed everywhere, future PRs will handle them.

To test, add param to some tests choosing proto version; we want to test
both 2 and 3 until we fully migrate.

ref https://github.com/neondatabase/neon/issues/10326

---------

Co-authored-by: Arthur Petukhovsky <petuhovskiy@yandex.ru>

test_runner/regress/test_normal_work.py

@@ -6,9 +6,14 @@ from fixtures.neon_fixtures import NeonEnv, NeonEnvBuilder
 from fixtures.pageserver.http import PageserverHttpClient
 
 
-def check_tenant(env: NeonEnv, pageserver_http: PageserverHttpClient):
+def check_tenant(
+    env: NeonEnv, pageserver_http: PageserverHttpClient, safekeeper_proto_version: int
+):
     tenant_id, timeline_id = env.create_tenant()
-    endpoint = env.endpoints.create_start("main", tenant_id=tenant_id)
+    config_lines = [
+        f"neon.safekeeper_proto_version = {safekeeper_proto_version}",
+    ]
+    endpoint = env.endpoints.create_start("main", tenant_id=tenant_id, config_lines=config_lines)
     # we rely upon autocommit after each statement
     res_1 = endpoint.safe_psql_many(
         queries=[
@@ -33,7 +38,14 @@ def check_tenant(env: NeonEnv, pageserver_http: PageserverHttpClient):
 
 
 @pytest.mark.parametrize("num_timelines,num_safekeepers", [(3, 1)])
-def test_normal_work(neon_env_builder: NeonEnvBuilder, num_timelines: int, num_safekeepers: int):
+# Test both proto versions until we fully migrate.
+@pytest.mark.parametrize("safekeeper_proto_version", [2, 3])
+def test_normal_work(
+    neon_env_builder: NeonEnvBuilder,
+    num_timelines: int,
+    num_safekeepers: int,
+    safekeeper_proto_version: int,
+):
     """
     Basic test:
     * create new tenant with a timeline
@@ -52,4 +64,4 @@ def test_normal_work(neon_env_builder: NeonEnvBuilder, num_timelines: int, num_s
     pageserver_http = env.pageserver.http_client()
 
     for _ in range(num_timelines):
-        check_tenant(env, pageserver_http)
+        check_tenant(env, pageserver_http, safekeeper_proto_version)

test_runner/regress/test_wal_acceptor_async.py

@@ -539,13 +539,16 @@ def test_recovery_uncommitted(neon_env_builder: NeonEnvBuilder):
     asyncio.run(run_recovery_uncommitted(env))
 
 
-async def run_wal_truncation(env: NeonEnv):
+async def run_wal_truncation(env: NeonEnv, safekeeper_proto_version: int):
     tenant_id = env.initial_tenant
     timeline_id = env.initial_timeline
 
     (sk1, sk2, sk3) = env.safekeepers
 
-    ep = env.endpoints.create_start("main")
+    config_lines = [
+        f"neon.safekeeper_proto_version = {safekeeper_proto_version}",
+    ]
+    ep = env.endpoints.create_start("main", config_lines=config_lines)
     ep.safe_psql("create table t (key int, value text)")
     ep.safe_psql("insert into t select generate_series(1, 100), 'payload'")
 
@@ -572,6 +575,7 @@ async def run_wal_truncation(env: NeonEnv):
     sk2.start()
     ep = env.endpoints.create_start(
         "main",
+        config_lines=config_lines,
     )
     ep.safe_psql("insert into t select generate_series(1, 200), 'payload'")
 
@@ -590,11 +594,13 @@ async def run_wal_truncation(env: NeonEnv):
 
 # Simple deterministic test creating tail of WAL on safekeeper which is
 # truncated when majority without this sk elects walproposer starting earlier.
-def test_wal_truncation(neon_env_builder: NeonEnvBuilder):
+# Test both proto versions until we fully migrate.
+@pytest.mark.parametrize("safekeeper_proto_version", [2, 3])
+def test_wal_truncation(neon_env_builder: NeonEnvBuilder, safekeeper_proto_version: int):
     neon_env_builder.num_safekeepers = 3
     env = neon_env_builder.init_start()
 
-    asyncio.run(run_wal_truncation(env))
+    asyncio.run(run_wal_truncation(env, safekeeper_proto_version))
 
 
 async def run_segment_init_failure(env: NeonEnv):