Details
-
Bug
-
Status: In Progress (View Workflow)
-
Critical
-
Resolution: Unresolved
-
10.11, 11.4
-
None
Description
According to thiru,
https://github.com/MariaDB/server/commit/3cef4f8f0fc88ae5bfae4603d8d600ec84cc70a9
optimization, with "single undo log record TRX_UNDO_EMPTY" does not work
It does not work if I LOCK TABLE WRITE prior to LOAD DATA INFILE, it does not work if I start transaction after LOCK TABLE WRITE, and it does not work, if I also add "set autocommit=0" to the mix.
Loading 1,4G worth of data into empty table, with any of the above tricks, profiler shows 15% of the time spent on row_purge_reset_trx_id()
Can't enable "bulk optimization" because of the bug MDEV-34703 I filed previously,
can't increase innodb sort buffer sizes, because actual requirement for tens of GB worth data
is unknown.
Attachments
Issue Links
- blocks
-
MDEV-33188 Enhance mariadb-dump and mariadb-import capabilities similar to MyDumper
- Stalled
- relates to
-
MDEV-515 innodb bulk insert
- Closed
-
MDEV-5171 Add support for --innodb-optimize-keys to mysqldump.
- Closed
-
MDEV-24621 In bulk insert, pre-sort and build indexes one page at a time
- Closed
-
MDEV-24813 Locking full table scan fails to use table-level locking
- In Review
-
MDEV-33625 Add option --dir to mariadb-dump
- Closed
-
MDEV-34703 Innodb bulk load : high IO, crashes on Linux, OOM on Linux, as tested with LOAD DATA INFILE
- Open