mysqld crashes when executing mysqldump
Description
Environment
I have reproduced this on an Ubuntu 16.04, using the official Percona repositories.
Attachments
Smart Checklist
Activity

KennT February 7, 2019 at 12:26 AM
commit 465a493741b838da4b93e554cf6108e4cd0f7a45 (origin/8.0-pxc-2345, 8.0-pxc-2345)
Merge: eced4c4c16b dcd2728b2cf
Author: Kenn Takara <kenn.takara@percona.com>
Date: Tue Jan 22 02:14:37 2019 -1000
Merge remote-tracking branch 'origin/5.7-pxc-2345' into 8.0-pxc-2345
commit dcd2728b2cfc86e583bd10c7fb6b30999a4dcb8a (origin/5.7-pxc-2345)
Author: Kenn Takara <kenn.takara@percona.com>
Date: Mon Jan 21 02:49:14 2019 -0800
: Crash triggered with mysqldump single-transaction
Issue
Running 'mysqldump --single-transaction' on a running PXC
server causes a crash.
Solution
If WSREP is enabled, skip over the consistent snapshots.
This was regression from the previous version.

Thomas Lutz January 17, 2019 at 10:59 AM
Since enabling binary logging seems to be a good idea in general, I tried this and it works. Thanks everybody! Hope this bug gets fixed soon.

Lalit Choudhary January 17, 2019 at 10:47 AM
Hi
Issue Consistently reproducible with Percona XtraDB Cluster when log_bin=OFF.
As a workaround, you can enable binary logs( log_bin=ON )and expire_log_days=1(will purge binary log older than 1 day) to avoid disk space issue.

NOC Sollie January 17, 2019 at 10:41 AM
@Thomas Lutz We did it by manually removing Percona and then installing the previous version from the .deb packages (download here)

Pierre BOESCH January 17, 2019 at 10:40 AM
Hi, percona does not provide old version in their repo. The solution is to use local apt archives:
```
cd /var/cache/apt/archives/
dpkg -i percona-xtradb-cluster-57_5.7.23-31.31-2.jessie_amd64.deb percona-xtradb-cluster-server-5.7_5.7.23-31.31-2.jessie_amd64.deb percona-xtradb-cluster-common-5.7_5.7.23-31.31-2.jessie_amd64.deb percona-xtradb-cluster-client-5.7_5.7.23-31.31-2.jessie_amd64.deb
```
I began by node3, node2 and finally node1. No down during the operation. Don't forget to do `mysql_ugrade` after.
Details
Details
Assignee

Reporter

Fix versions
Affects versions
Priority
Smart Checklist
Open Smart Checklist
Smart Checklist

I have a Galera cluster setup, with the following versions of packages:
And when i tried to make a backup i got this:
in mysqld error log i got this:
I have also reproduced this with GDB, i have attached the backtrace and my my.cnf as for reference.