From: Vladimir Davydov <vdavydov.dev@gmail.com>
To: kostja@tarantool.org
Cc: tarantool-patches@freelists.org
Subject: [PATCH 0/4] Fix DML vs DDL race
Date: Sun, 24 Mar 2019 00:07:20 +0300 [thread overview]
Message-ID: <cover.1553373792.git.vdavydov.dev@gmail.com> (raw)
In Vinyl a DML request may yield on disk read while still having a
pointer to the target space saved on stack. If the space is altered
during this time window, the DML request will most certainly crash
while trying to dereference the stale pointer when it resumes. The
solution is simple: abort all affected transactions whenever we delete
a space from the cache.
https://github.com/tarantool/tarantool/issues/3420
https://github.com/tarantool/tarantool/commits/dv/gh-3420-vy-abort-tx-in-ddl
NOTE the branch is based on top of
https://github.com/tarantool/tarantool/commits/dv/gh-4070-vy-fail-aborted-transactions-early
which was submitted for review earlier:
https://www.freelists.org/post/tarantool-patches/PATCH-12-vinyl-drop-useless-vy-read-viewis-aborted-flag
https://www.freelists.org/post/tarantool-patches/PATCH-22-vinyl-fail-aborted-transactions-early
Vladimir Davydov (4):
vinyl: introduce hash of spaces affected by transaction
vinyl: don't abort transactions that modify only local spaces for ro
vinyl: abort affected transactions when space is removed from cache
Revert "test: skip ddl test for vinyl on travis"
rpm/tarantool.spec | 2 -
src/box/blackhole.c | 1 +
src/box/memtx_space.c | 1 +
src/box/schema.cc | 3 ++
src/box/space.c | 6 +++
src/box/space.h | 15 ++++++
src/box/sysview.c | 1 +
src/box/vinyl.c | 33 ++++++++++---
src/box/vy_point_lookup.c | 14 ++++++
src/box/vy_read_iterator.c | 13 +++++
src/box/vy_tx.c | 45 ++++++++++++++---
src/box/vy_tx.h | 44 +++++++++++++++--
test/vinyl/ddl.skipcond | 6 ---
test/vinyl/errinj_ddl.result | 110 +++++++++++++++++++++++++++++++++++++++++
test/vinyl/errinj_ddl.test.lua | 51 +++++++++++++++++++
test/vinyl/misc.result | 38 ++++++++++++++
test/vinyl/misc.test.lua | 16 ++++++
17 files changed, 376 insertions(+), 23 deletions(-)
delete mode 100644 test/vinyl/ddl.skipcond
--
2.11.0
next reply other threads:[~2019-03-23 21:07 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-03-23 21:07 Vladimir Davydov [this message]
2019-03-23 21:07 ` [PATCH 1/4] vinyl: introduce hash of spaces affected by transaction Vladimir Davydov
2019-03-28 13:25 ` [tarantool-patches] " Konstantin Osipov
2019-03-28 13:58 ` Vladimir Davydov
2019-03-23 21:07 ` [PATCH 2/4] vinyl: don't abort transactions that modify only local spaces for ro Vladimir Davydov
2019-03-25 5:27 ` [tarantool-patches] " Георгий Кириченко
2019-03-25 8:13 ` Vladimir Davydov
2019-03-25 8:58 ` Георгий Кириченко
2019-03-25 9:30 ` Vladimir Davydov
2019-03-23 21:07 ` [PATCH 3/4] vinyl: abort affected transactions when space is removed from cache Vladimir Davydov
2019-03-25 5:26 ` [tarantool-patches] " Георгий Кириченко
2019-03-25 8:17 ` Vladimir Davydov
[not found] ` <1564677.EMV258VVK2@home.lan>
2019-03-25 9:51 ` Vladimir Davydov
2019-03-25 5:45 ` Георгий Кириченко
2019-03-25 8:21 ` Vladimir Davydov
2019-03-25 9:03 ` Георгий Кириченко
2019-03-28 13:45 ` [tarantool-patches] " Konstantin Osipov
2019-03-28 14:02 ` Vladimir Davydov
2019-03-28 14:12 ` Konstantin Osipov
2019-03-23 21:07 ` [PATCH 4/4] Revert "test: skip ddl test for vinyl on travis" Vladimir Davydov
2019-03-28 13:46 ` [tarantool-patches] " Konstantin Osipov
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=cover.1553373792.git.vdavydov.dev@gmail.com \
--to=vdavydov.dev@gmail.com \
--cc=kostja@tarantool.org \
--cc=tarantool-patches@freelists.org \
--subject='Re: [PATCH 0/4] Fix DML vs DDL race' \
/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