From: Vladimir Davydov <vdavydov.dev@gmail.com>
To: Serge Petrenko <sergepetrenko@tarantool.org>
Cc: tarantool-patches@freelists.org,
Georgy Kirichenko <georgy@tarantool.org>
Subject: Re: [PATCH] replication: move cluster id match check to replica
Date: Thu, 7 Feb 2019 13:14:51 +0300 [thread overview]
Message-ID: <20190207101451.hajah3gf2woe6qr7@esperanza> (raw)
In-Reply-To: <330D09EB-772C-4943-A9C0-77BF4A1889F5@tarantool.org>
Pushed to 2.1 with some minor changes:
diff --git a/src/box/applier.cc b/src/box/applier.cc
index 045c6910..512d05df 100644
--- a/src/box/applier.cc
+++ b/src/box/applier.cc
@@ -410,7 +410,7 @@ 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
+ * 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.
*/
@@ -423,7 +423,7 @@ applier_subscribe(struct applier *applier)
* assume that it has done all the checks.
* In this case cluster_id will remain zero.
*/
- if (!tt_uuid_is_equal(&cluster_id, &uuid_nil) &&
+ 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),
diff --git a/src/box/box.cc b/src/box/box.cc
index 8a8c0816..e12a1cba 100644
--- a/src/box/box.cc
+++ b/src/box/box.cc
@@ -1609,11 +1609,11 @@ box_process_subscribe(struct ev_io *io, struct xrow_header *header)
* the replica how many rows we have in stock for it,
* and identify ourselves with our own replica id.
*
- * Since Tarantool version 2.1.1 master doesn't check
- * that replica and master have the same cluster id.
- * It sends its cluster id to replica instead, and replica
- * checks that its cluster id matches master's one.
- * Older versions will just ignore the additional field.
+ * 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_subscribe_response_xc(&row,
diff --git a/test/replication/misc.result b/test/replication/misc.result
index 2401ac5c..ab827c50 100644
--- a/test/replication/misc.result
+++ b/test/replication/misc.result
@@ -634,7 +634,7 @@ 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.
+-- master believes replica is in cluster, but their cluster UUIDs differ.
test_run:cmd("start server replica")
---
- true
@@ -647,7 +647,6 @@ box.info.replication[2].downstream.status
---
- stopped
...
-test_run:cmd("restart server default with cleanup=1")
test_run:cmd("stop server replica")
---
- true
@@ -660,3 +659,9 @@ 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 a1ddf75a..eda5310b 100644
--- a/test/replication/misc.test.lua
+++ b/test/replication/misc.test.lua
@@ -255,12 +255,13 @@ 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.
+-- 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("restart server default with cleanup=1")
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')
prev parent reply other threads:[~2019-02-07 10:14 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-01-31 13:25 Serge Petrenko
2019-01-31 13:30 ` Serge Petrenko
2019-02-06 10:48 ` Vladimir Davydov
2019-02-06 14:51 ` Serge Petrenko
2019-02-07 10:14 ` Vladimir Davydov [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190207101451.hajah3gf2woe6qr7@esperanza \
--to=vdavydov.dev@gmail.com \
--cc=georgy@tarantool.org \
--cc=sergepetrenko@tarantool.org \
--cc=tarantool-patches@freelists.org \
--subject='Re: [PATCH] replication: move cluster id match check to replica' \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox