[tarantool-patches] Re: [PATCH v2 0/4] In-memory wal replication
v.shpilevoy at tarantool.org
Thu Sep 19 00:45:43 MSK 2019
Hi! Thanks for the patchset!
[NOTE: review is not finished yet, I will send more emails]
Please, don't omit branch and issue links. You have 4 branches
having 3794 in their name. You could save a little bit of
reviewer's time by specifying a branch here.
Looks like your branch fails on travis. Please, fix it.
On 18/09/2019 11:36, Georgy Kirichenko wrote:
> This patchset introduces from-wal-memory-replication and consist of 4
> The first patch purges distinct relay facilities to advanced gc consumer
> state. Also it changes gc behaviour - now gc compares consumers using
> lsns corresponding to local instance id instead of vclock signature.
> Howevers this changes gc behavior, its look like we could restore the
> gc behavior after relay would be migrated to the wal thread. The whole
> patch is essential because in relay switched to memory relaying there is
> no more file boundaries and on_close_log trigger invocation now.
So you've broken GC and promise to fix it or what? Not sure if I
> The second patch introduces xrow buffer structure to be used by wal
> while logging. Wal encodes rows to a memory buffer (xrow_buffer
> structure) and then writes encoded data to a xlog file.
> The next patch implements xrow_buffer_cursor to have an ability to found a
> position in the buffer by vclock and then fetch all rows one-by-one
> until the last row.
> The last patch makes a relay able to read from wal memory using
> a xrow_buffer cursor. When relay went out of wal memory window it turns
> back to file mode. After last known xlog file is relayed the a relay
> returns to wal memory.
Please, ask Alexander to bench your branch to check if there are no
> Changes in v2:
> - Get rid of gc refactoring
> - Rename wal_memory to xrow_buffer according to review
> - Commentaries
> - Bug fixes
> Georgy Kirichenko (4):
> relay: adjust gc state on relay status update
> wal: wal memory buffer
> wal: xrow buffer cursor
> replication: use wal memory buffer to fetch rows
> src/box/CMakeLists.txt | 1 +
> src/box/gc.c | 7 +-
> src/box/relay.cc | 275 ++++++-----------
> src/box/wal.c | 193 +++++++++++-
> src/box/wal.h | 36 +++
> src/box/xlog.c | 77 +++--
> src/box/xlog.h | 9 +
> src/box/xrow_buf.c | 279 ++++++++++++++++++
> src/box/xrow_buf.h | 160 ++++++++++
> src/lib/core/cbus.c | 4 +
> src/lib/core/errinj.h | 1 +
> test/box/errinj.result | 2 +
> test/replication/force_recovery.result | 8 +
> test/replication/force_recovery.test.lua | 2 +
> test/replication/replica_rejoin.result | 8 +
> test/replication/replica_rejoin.test.lua | 2 +
> .../show_error_on_disconnect.result | 8 +
> .../show_error_on_disconnect.test.lua | 2 +
> test/replication/suite.ini | 2 +-
> test/xlog/panic_on_wal_error.result | 12 +
> test/xlog/panic_on_wal_error.test.lua | 3 +
> test/xlog/suite.ini | 2 +-
> 22 files changed, 874 insertions(+), 219 deletions(-)
> create mode 100644 src/box/xrow_buf.c
> create mode 100644 src/box/xrow_buf.h
More information about the Tarantool-patches