Semi-Sync buffer overflow

[Robert Navarro]

mysql> \s
--------------
mysql Ver 14.14 Distrib 5.7.15-9, for debian-linux-gnu (x86_64) using 6.3

Connection id: 38743
Current database:
Current user: root@localhost
SSL: Not in use
Current pager: stdout
Using outfile: ''
Using delimiter: ;
Server version: 5.7.15-9-log Percona Server (GPL), Release '9', Revision '9f0fd0a'
Protocol version: 10
Connection: Localhost via UNIX socket
Server characterset: utf8
Db characterset: utf8
Client characterset: latin1
Conn. characterset: latin1
UNIX socket: /vt/vtdataroot/vt_0000003601/mysql.sock
Uptime: 1 day 17 hours 57 min 33 sec

root@vttablet-3601-2663081908-3prle:/# cat /etc/*-release
PRETTY_NAME="Debian GNU/Linux 8 (jessie)"
NAME="Debian GNU/Linux"
VERSION_ID="8"
VERSION="8 (jessie)"
ID=debian
HOME_URL="http://www.debian.org/"
SUPPORT_URL="http://www.debian.org/support"
BUG_REPORT_URL="https://bugs.debian.org/"

I have attached the dump we get from the MySQL process to this bug.

This usually happens when we're changing around slaves that are attached to the master. We'll turn on replication for a new slave and that'll cause this crash.

Let me know if you need any other information about this, I'm happy to help get it. This is really high priority for us and we would rather not have to turn off Semi-Sync for data integrity purposes.

[Robert Navarro (crshman)]

This is a pretty pressing issue for us, each time we re-parent our shards there is a non-trivial percent chance of the master crashing during the process.

What other information is needed to help move this along?

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PS-3598