diff --git a/src/box/applier.cc b/src/box/applier.cc
index 21d2e6bcb90066d0f510641489ab9d1130bfae88..512d05dfa1df4149d65fa141fed94126cf650231 100644
--- a/src/box/applier.cc
+++ b/src/box/applier.cc
@@ -391,6 +391,7 @@ applier_subscribe(struct applier *applier)
struct ibuf *ibuf = &applier->ibuf;
struct xrow_header row;
struct vclock remote_vclock_at_subscribe;
+ struct tt_uuid cluster_id = uuid_nil;
xrow_encode_subscribe_xc(&row, &REPLICASET_UUID, &INSTANCE_UUID,
&replicaset.vclock);
@@ -408,9 +409,26 @@ applier_subscribe(struct applier *applier)
/*
* In case of successful subscribe, the server
* responds with its current vclock.
+ *
+ * Tarantool > 2.1.1 also sends its cluster id to
+ * the replica, and replica has to check whether
+ * its and master's cluster ids match.
*/
vclock_create(&remote_vclock_at_subscribe);
- xrow_decode_vclock_xc(&row, &remote_vclock_at_subscribe);
+ xrow_decode_subscribe_response_xc(&row,
+ &cluster_id,
+ &remote_vclock_at_subscribe);
+ /*
+ * If master didn't send us its cluster id
+ * assume that it has done all the checks.
+ * In this case cluster_id will remain zero.
+ */
+ if (!tt_uuid_is_nil(&cluster_id) &&
+ !tt_uuid_is_equal(&cluster_id, &REPLICASET_UUID)) {
+ tnt_raise(ClientError, ER_REPLICASET_UUID_MISMATCH,
+ tt_uuid_str(&cluster_id),
+ tt_uuid_str(&REPLICASET_UUID));
+ }
}
/*
* Tarantool < 1.6.7:
diff --git a/src/box/box.cc b/src/box/box.cc
index 8892d0f0e3e41f19b2650e92d71186b374ca79a3..e12a1cba423b6fbfd85be48159edc772054f9402 100644
--- a/src/box/box.cc
+++ b/src/box/box.cc
@@ -1584,18 +1584,6 @@ box_process_subscribe(struct ev_io *io, struct xrow_header *header)
/* Check permissions */
access_check_universe_xc(PRIV_R);
- /**
- * Check that the given UUID matches the UUID of the
- * replica set this replica belongs to. Used to handshake
- * replica connect, and refuse a connection from a replica
- * which belongs to a different replica set.
- */
- if (!tt_uuid_is_equal(&replicaset_uuid, &REPLICASET_UUID)) {
- tnt_raise(ClientError, ER_REPLICASET_UUID_MISMATCH,
- tt_uuid_str(&REPLICASET_UUID),
- tt_uuid_str(&replicaset_uuid));
- }
-
/* Check replica uuid */
struct replica *replica = replica_by_uuid(&replica_uuid);
if (replica == NULL || replica->id == REPLICA_ID_NIL) {
@@ -1620,9 +1608,17 @@ box_process_subscribe(struct ev_io *io, struct xrow_header *header)
* Send a response to SUBSCRIBE request, tell
* the replica how many rows we have in stock for it,
* and identify ourselves with our own replica id.
+ *
+ * Tarantool > 2.1.1 master doesn't check that replica
+ * has the same cluster id. Instead it sends its cluster
+ * id to replica, and replica checks that its cluster id
+ * matches master's one. Older versions will just ignore
+ * the additional field.
*/
struct xrow_header row;
- xrow_encode_vclock_xc(&row, &replicaset.vclock);
+ xrow_encode_subscribe_response_xc(&row,
+ &REPLICASET_UUID,
+ &replicaset.vclock);
/*
* Identify the message with the replica id of this
* instance, this is the only way for a replica to find
diff --git a/src/box/xrow.c b/src/box/xrow.c
index c4e3073be7917b41ebe6d086539027fb30152c57..ba3671fed819202fa208940589b1ecac5c19420e 100644
--- a/src/box/xrow.c
+++ b/src/box/xrow.c
@@ -1170,6 +1170,35 @@ xrow_encode_vclock(struct xrow_header *row, const struct vclock *vclock)
return 0;
}
+int
+xrow_encode_subscribe_response(struct xrow_header *row,
+ const struct tt_uuid *replicaset_uuid,
+ const struct vclock *vclock)
+{
+ memset(row, 0, sizeof(*row));
+ size_t size = mp_sizeof_map(2) +
+ mp_sizeof_uint(IPROTO_VCLOCK) + mp_sizeof_vclock(vclock) +
+ mp_sizeof_uint(IPROTO_CLUSTER_UUID) +
+ mp_sizeof_str(UUID_STR_LEN);
+ char *buf = (char *) region_alloc(&fiber()->gc, size);
+ if (buf == NULL) {
+ diag_set(OutOfMemory, size, "region_alloc", "buf");
+ return -1;
+ }
+ char *data = buf;
+ data = mp_encode_map(data, 2);
+ data = mp_encode_uint(data, IPROTO_VCLOCK);
+ data = mp_encode_vclock(data, vclock);
+ data = mp_encode_uint(data, IPROTO_CLUSTER_UUID);
+ data = xrow_encode_uuid(data, replicaset_uuid);
+ assert(data <= buf + size);
+ row->body[0].iov_base = buf;
+ row->body[0].iov_len = (data - buf);
+ row->bodycnt = 1;
+ row->type = IPROTO_OK;
+ return 0;
+}
+
void
xrow_encode_timestamp(struct xrow_header *row, uint32_t replica_id, double tm)
{
diff --git a/src/box/xrow.h b/src/box/xrow.h
index 2654e35e6355de7113f02f285da7567cb1e56dbd..719add4f0df7dcad857ad70d193baa864ec45414 100644
--- a/src/box/xrow.h
+++ b/src/box/xrow.h
@@ -348,6 +348,37 @@ xrow_decode_vclock(struct xrow_header *row, struct vclock *vclock)
return xrow_decode_subscribe(row, NULL, NULL, vclock, NULL);
}
+/**
+ * Encode a response to subscribe request.
+ * @param row[out] Row to encode into.
+ * @param replicaset_uuid.
+ * @param vclock.
+ *
+ * @retval 0 Success.
+ * @retval -1 Memory error.
+ */
+int
+xrow_encode_subscribe_response(struct xrow_header *row,
+ const struct tt_uuid *replicaset_uuid,
+ const struct vclock *vclock);
+
+/**
+ * Decode a response to subscribe request.
+ * @param row Row to decode.
+ * @param[out] replicaset_uuid.
+ * @param[out] vclock.
+ *
+ * @retval 0 Success.
+ * @retval -1 Memory or format error.
+ */
+static inline int
+xrow_decode_subscribe_response(struct xrow_header *row,
+ struct tt_uuid *replicaset_uuid,
+ struct vclock *vclock)
+{
+ return xrow_decode_subscribe(row, replicaset_uuid, NULL, vclock, NULL);
+}
+
/**
* Encode a heartbeat message.
* @param row[out] Row to encode into.
@@ -759,6 +790,26 @@ xrow_decode_vclock_xc(struct xrow_header *row, struct vclock *vclock)
diag_raise();
}
+/** @copydoc xrow_encode_subscribe_response. */
+static inline void
+xrow_encode_subscribe_response_xc(struct xrow_header *row,
+ const struct tt_uuid *replicaset_uuid,
+ const struct vclock *vclock)
+{
+ if (xrow_encode_subscribe_response(row, replicaset_uuid, vclock) != 0)
+ diag_raise();
+}
+
+/** @copydoc xrow_decode_subscribe_response. */
+static inline void
+xrow_decode_subscribe_response_xc(struct xrow_header *row,
+ struct tt_uuid *replicaset_uuid,
+ struct vclock *vclock)
+{
+ if (xrow_decode_subscribe_response(row, replicaset_uuid, vclock) != 0)
+ diag_raise();
+}
+
/** @copydoc iproto_reply_ok. */
static inline void
iproto_reply_ok_xc(struct obuf *out, uint64_t sync, uint32_t schema_version)
diff --git a/test/replication/misc.result b/test/replication/misc.result
index c32681a7ada4c333c8bd2e7ad1ed94c1283d7b9c..ab827c50135cf328842c7e1ae1c8021489ef3e67 100644
--- a/test/replication/misc.result
+++ b/test/replication/misc.result
@@ -604,3 +604,64 @@ test_run:cmd("delete server replica")
test_run:cleanup_cluster()
---
...
+--
+-- gh-3704 move cluster id check to replica
+--
+test_run:cmd("create server replica with rpl_master=default, script='replication/replica.lua'")
+---
+- true
+...
+box.schema.user.grant("guest", "replication")
+---
+...
+test_run:cmd("start server replica")
+---
+- true
+...
+test_run:grep_log("replica", "REPLICASET_UUID_MISMATCH")
+---
+- null
+...
+box.info.replication[2].downstream.status
+---
+- follow
+...
+-- change master's cluster uuid and check that replica doesn't connect.
+test_run:cmd("stop server replica")
+---
+- true
+...
+_ = box.space._schema:replace{'cluster', tostring(uuid.new())}
+---
+...
+-- master believes replica is in cluster, but their cluster UUIDs differ.
+test_run:cmd("start server replica")
+---
+- true
+...
+test_run:wait_log("replica", "REPLICASET_UUID_MISMATCH", nil, 1.0)
+---
+- REPLICASET_UUID_MISMATCH
+...
+box.info.replication[2].downstream.status
+---
+- stopped
+...
+test_run:cmd("stop server replica")
+---
+- true
+...
+test_run:cmd("cleanup server replica")
+---
+- true
+...
+test_run:cmd("delete server replica")
+---
+- true
+...
+test_run:cleanup_cluster()
+---
+...
+box.schema.user.revoke('guest', 'replication')
+---
+...
diff --git a/test/replication/misc.test.lua b/test/replication/misc.test.lua
index 6a8af05c37661db5a9533d32c9bb7945d256c693..eda5310b6588a78044e94ba063781ec40f141ca9 100644
--- a/test/replication/misc.test.lua
+++ b/test/replication/misc.test.lua
@@ -243,3 +243,25 @@ test_run:cmd("stop server replica")
test_run:cmd("cleanup server replica")
test_run:cmd("delete server replica")
test_run:cleanup_cluster()
+
+--
+-- gh-3704 move cluster id check to replica
+--
+test_run:cmd("create server replica with rpl_master=default, script='replication/replica.lua'")
+box.schema.user.grant("guest", "replication")
+test_run:cmd("start server replica")
+test_run:grep_log("replica", "REPLICASET_UUID_MISMATCH")
+box.info.replication[2].downstream.status
+-- change master's cluster uuid and check that replica doesn't connect.
+test_run:cmd("stop server replica")
+_ = box.space._schema:replace{'cluster', tostring(uuid.new())}
+-- master believes replica is in cluster, but their cluster UUIDs differ.
+test_run:cmd("start server replica")
+test_run:wait_log("replica", "REPLICASET_UUID_MISMATCH", nil, 1.0)
+box.info.replication[2].downstream.status
+
+test_run:cmd("stop server replica")
+test_run:cmd("cleanup server replica")
+test_run:cmd("delete server replica")
+test_run:cleanup_cluster()
+box.schema.user.revoke('guest', 'replication')