Tarantool development patches archive
 help / color / mirror / Atom feed
* [Tarantool-patches] [PATCH v8 0/2] filter incoming packets
@ 2021-06-11 15:22 Cyrill Gorcunov via Tarantool-patches
  2021-06-11 15:22 ` [Tarantool-patches] [PATCH v8 1/2] applier: filter incoming synchro packets via transaction initiator Cyrill Gorcunov via Tarantool-patches
  2021-06-11 15:22 ` [Tarantool-patches] [PATCH v8 2/2] Vlad: applier filtration Cyrill Gorcunov via Tarantool-patches
  0 siblings, 2 replies; 6+ messages in thread
From: Cyrill Gorcunov via Tarantool-patches @ 2021-06-11 15:22 UTC (permalink / raw)
  To: tml; +Cc: Vladislav Shpilevoy

Guys, here is a patch which implements the filtration we are
interested in. The patch is not for merging yet because there
some weird things going on:

 - to verify that we're hitting exactly the issue we're interested
   in I apply Vlad's patch to make sure that replication doesn't
   pass where is should

 - the test sets up somehow weird configuration but changing it
   even a bit doesnt trigger the problem we're investigating,
   and to be honest I'm not sure why

 - current test (with Vlad's patch applied) returns
	box.space.sync:select{}
	 | ---
	 | - []
	 | ...

   in turn if we revert Vlad's patch then data get propagated
   to replica1 as it should, which proves that we need to
   filter by xrow->replica_id

Still I would like to get a feedback how to simplify the testcase.

branch gorcunov/gh-6035-applier-filter-7-notest

Cyrill Gorcunov (2):
  applier: filter incoming synchro packets via transaction initiator
  Vlad: applier filtration

 src/box/applier.cc                            |  29 ++--
 src/lib/raft/raft.h                           |   7 -
 .../gh-5445-leader-inconsistency.result       |  15 ++
 .../gh-5445-leader-inconsistency.test.lua     |   5 +
 .../replication/gh-6035-applier-filter.result | 137 ++++++++++++++++++
 .../gh-6035-applier-filter.test.lua           |  64 ++++++++
 test/replication/gh-6035-master.lua           |   1 +
 test/replication/gh-6035-node.lua             |  35 +++++
 test/replication/gh-6035-replica1.lua         |   1 +
 test/replication/gh-6035-replica2.lua         |   1 +
 test/replication/suite.cfg                    |   3 +
 11 files changed, 278 insertions(+), 20 deletions(-)
 create mode 100644 test/replication/gh-6035-applier-filter.result
 create mode 100644 test/replication/gh-6035-applier-filter.test.lua
 create mode 120000 test/replication/gh-6035-master.lua
 create mode 100644 test/replication/gh-6035-node.lua
 create mode 120000 test/replication/gh-6035-replica1.lua
 create mode 120000 test/replication/gh-6035-replica2.lua


base-commit: 9fb95cfd78403b33cacbe765c35036a0532b2c02
-- 
2.31.1


^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2021-06-15 11:55 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-06-11 15:22 [Tarantool-patches] [PATCH v8 0/2] filter incoming packets Cyrill Gorcunov via Tarantool-patches
2021-06-11 15:22 ` [Tarantool-patches] [PATCH v8 1/2] applier: filter incoming synchro packets via transaction initiator Cyrill Gorcunov via Tarantool-patches
2021-06-15 10:36   ` Serge Petrenko via Tarantool-patches
2021-06-15 11:35   ` Serge Petrenko via Tarantool-patches
2021-06-15 11:55     ` Serge Petrenko via Tarantool-patches
2021-06-11 15:22 ` [Tarantool-patches] [PATCH v8 2/2] Vlad: applier filtration Cyrill Gorcunov via Tarantool-patches

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox