Issues

Upstream: OFFSET 99 LIMIT 1 on a compound PK takes ages to execute.

Commits stalled behind publish_coordinates_for_global_status mutexes

Adopt InnoDB AIO layer to io_uring interface

Implement TC_NULLBINLOG transaction coordinator

innodb_empty_free_list_algorithm incorrect default

Remove AHI entries lazily after a table or index is dropped

Remove AHI overhead for DDLs when AHI is disabled

Slow shutdown with many tables

Performance degradation during sustained full table scan

InnoDB's MVCC has O(N^2) behaviors

4000x query performance regression in 8.0

Enable hardware CRC32 under Valgrind

Slow startup after an upgrade from MySQL 5.7 to MySQL 8.

Severe degradation of Innodb buffer pool load times at startup after migration from 5.5. to 5.6

Restore sched_getcpu fuzzy counters in InnoDB

LP #1742751: mysqldump takes 5x as long with 5.7.20-19

LP #1740459: replication delay too much on tokudb partition table

LP #1728512: A follow-up fix for buffer pool mutex split patch might be suboptimal, commit 2

LP #1728509: A follow-up fix for buffer pool mutex split patch might be suboptimal, commit 1

LP #1716817: HIGH CPU usage

Unnecessary flush list iteration during ALTER TABLE ADD INDEX

LP #1690399: InnoDB performance drop in 5.7 because of the lru_manager

LP #1599520: "show warnings" uses an abnormal amount of system CPU if client compression is ON

LP #1589334: After upgrading to 5.7 mysqldump much slower

LP #1534114: Busy server prefers LRU flushing over flush list flushing too strongly

LP #1532747: Page cleaner worker threads are not instrumented for performance schema

LP #1525215: buf_pool_resize can lock less in checking whether AHI is on or off

LP #1298812: Performance regression on long-running RW workloads with InnoDB compression

LP #1131426: ORDERBY + LIMIT index not chosen when joining an extra table

LP #747112: kernel_mutex optimization around trx_create() and trx_free()

LP #1704267: Spurious lock_wait_timeout_thread wakeup in lock_wait_suspend_thread()

LP #1704195: Release lock on table statistics after query plan created | Stop ANALYZE TABLE from flushing table definition cache

LP #1679155: InnoDB adaptive hash index uses a bad partitioning algorithm for the real world

LP #1674338: Extend index-only prefix index query optimisation to secondary indexes on generated columns

LP #1674337: Extend index-only prefix index query optimisation to JSON

LP #1664891: Investigate whether buf_page_set_accessed locking can be weakened

LP #1618393: Optimize ibuf merge when reading a page from disk

LP #1612650: audit_log plugin contention on libc TZ conversion

LP #1411694: btr_search_guess_on_hash makes found block young twice?

LP #1374930: put threads with GLOBAL READ lock (executing ftwrl) into high priority queue

LP #1369373: minor changes to optimize innodb_kill_idle_transaction

LP #1237702: InnoDB adaptive flushing: Incorrect calculation of the lsn_limit

LP #1188219: ineffective searching for unreserved slot in srv_conc_queue

LP #1185686: Replace InnoDB heap allocation with direct malloc() in trx_create()

LP #1108701: Sysbench performance regression due to commit 410 (bitmap)

LP #1073170: Fadvise the slave SQL thread for relay logs

LP #952920: performance tuning for deadlock detect switch

LP #690019: split_doublewrite for each write IO thread

LP #1264952: Adjust perfschema autosizing for Percona Server mutexes