Hi, Nikita! On Nov 24, Nikita Malyavin wrote:
On Mon, 15 Nov 2021 at 13:52, Sergei Golubchik <serg@mariadb.org> wrote:
Hi, Nikita!
On Nov 14, Nikita Malyavin wrote:
revision-id: a5fca9a6e30 (mariadb-10.5.2-478-ga5fca9a6e30) parent(s): ad6e7b87107 author: Nikita Malyavin committer: Nikita Malyavin timestamp: 2021-01-27 17:28:05 +1000 message:
MENT-651 [6/8] store cache managers in a list
again, a couple of lines with a description would be good here
I think it should be just a fixup for [4/8] ALTER ONLINE TABLE.
diff --git a/mysql-test/suite/binlog/t/online_alter.test b/mysql-test/suite/binlog/t/online_alter.test index 804e847d008..4ed2db74bd6 100644 --- a/mysql-test/suite/binlog/t/online_alter.test +++ b/mysql-test/suite/binlog/t/online_alter.test @@ -218,7 +219,7 @@ set autocommit = 0; start transaction; update t1 set b= 7007 where a = 7; --error ER_DUP_ENTRY -update t1 set a= 8 where a = 8 or a = 9; +update t1 set a= 8, b= 8008 where a = 8 or a = 9 order by a;
why did you change tests for what is, as far as I understand, just an optimization?
That's the point, that it's not!
The thing is, in case of an error, the table is invalidated. As far as I remember, it could happen even if the table could've just evicted from the cache (in case of a long transaction involving many tables).
So when committing, corresponding TABLE_SHARE may be inaccessible, by different reasons.
So it turns to practically correct that handlerton can't depend from TABLE or TABLE_SHARE.
Okay. Why did you have to change tests though?
commit; set autocommit = 1;
The rest of my questions... I now understand better what you're doing and why. There are some questions still, and I wanted to look in a debugger, but this branch doesn't compile with gcc 11.2, just too many errors. Could you please move your work on top of 10.8 please? Regards, Sergei VP of MariaDB Server Engineering and security@mariadb.org