[MDEV-21910] KIlling thread on Galera could cause mutex deadlock - Jira

XML

Word

Printable

Details

Type: Bug
Status: Closed (View Workflow)
Priority: Blocker
Resolution: Fixed
Affects Version/s: 10.1(EOL), 10.2(EOL), 10.3(EOL), 10.4(EOL), 10.5
Fix Version/s: 10.4.14, 10.5.5
Component/s: Galera, Storage Engine - InnoDB
Labels:
None

Description

There following issues here:

Whenever Galera BF (brute force) transaction decides to abort conflicting transaction it will kill that thread using thd::awake()
User KILL [QUERY|CONNECTION] ... for a thread it will also call thd::awake()
Whenever one of these actions is executed we will hold number of InnoDB internal mutexes and thd mutexes.
Sometimes these mutexes are taken in different order causing mutex deadlock.

Attachments

Attachments

- Sort By Name
- Sort By Date
- Ascending
- Descending
- Thumbnails
- List
- Download All

wsrep_kill.diff
6 kB
2020-07-28 14:21

Issue Links

causes

MDEV-23328 Server hang due to Galera lock conflict resolution

Closed

relates to

MDEV-18464 Port kill_one_trx fixes from 10.4 to 10.1

Closed

MDEV-23101 SIGSEGV in lock_rec_unlock() when Galera is enabled

Closed

MDEV-22666 galera.MW-328A MTR failed: "Semaphore wait has lasted > 600 seconds" and do not release port 16002

Closed

MDEV-23851 Galera assertion at lock0lock.cc line 655 because of BF-BF lock wait

Closed

Activity

People

Assignee:: Jan Lindström (Inactive)

Reporter:: Jan Lindström (Inactive)

Votes:: 1 Vote for this issue

Watchers:: 12 Start watching this issue

Dates

Created:: 2020-03-11 08:11

Updated:: 2024-07-07 22:14

Resolved:: 2020-07-29 10:48

Git Integration

Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.