I upgraded our 3-node galera cluster to 5.9.19-25-6 yesterday. The upgrade itself went flawlessly after pinning the repository to counter the Debian 7 version mixup. I ran a few tests afterwards and put the cluster back into production. Unfortunately, at about 5am monitoring alerts aborted my sleep, 2 out of 3 nodes were down. After I examined the logs, I figured both nodes crashed just after executing our backup routine.
I took one node to reproduce the problem. I can regularly crash a local node issuing this command (used to work for previous versions just fine):
mysqldump --verbose --single-transaction --routines --triggers --skip-lock-tables -u root testdb > test.sql
std err from mysqldump:
-- Connecting to localhost... -- Starting transaction... mysqldump: Couldn't execute 'START TRANSACTION /*!40100 WITH CONSISTENT SNAPSHOT */': Lost connection to MySQL server during query (2013)
node3.err log entry:
09:42:48 UTC - mysqld got signal 11 ; This could be because you hit a bug. It is also possible that this binary or one of the libraries it was linked against is corrupt, improperly built, or misconfigured. This error can also be caused by malfunctioning hardware. We will try our best to scrape up some info that will hopefully help diagnose the problem, but since we have already crashed, something is definitely wrong and this may fail. Please help us make Percona XtraDB Cluster better by reporting any bugs at https://bugs.launchpad.net/percona-xtradb-cluster key_buffer_size=0 read_buffer_size=2097152 max_used_connections=17 max_threads=602 thread_count=3 connection_count=1 It is possible that mysqld could use up to key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 1396126 K bytes of memory Hope that's ok; if not, decrease some variables in the equation. Thread pointer: 0x7fe16806d0a0 Attempting backtrace. You can use the following information to find out where mysqld died. If you see no messages after this, something went terribly wrong... stack_bottom = 7fe37806ee50 thread_stack 0x40000 /usr/sbin/mysqld(my_print_stacktrace+0x2e)[0x8c646e] /usr/sbin/mysqld(handle_fatal_signal+0x379)[0x67f7c9] /lib/x86_64-linux-gnu/libpthread.so.0(+0xf030)[0x7fe524b85030] /usr/sbin/mysqld[0x875a84] /usr/sbin/mysqld[0x5da17c] /usr/sbin/mysqld(_Z24plugin_foreach_with_maskP3THDPFcS0_P13st_plugin_intPvEijS3_+0x10d)[0x703bcd] /usr/sbin/mysqld(_Z28ha_start_consistent_snapshotP3THD+0x1cc)[0x5ddd9c] /usr/sbin/mysqld(_Z11trans_beginP3THDj+0x100)[0x7769b0] /usr/sbin/mysqld(_Z21mysql_execute_commandP3THD+0x1bbd)[0x6f5b2d] /usr/sbin/mysqld(_Z11mysql_parseP3THDPcjP12Parser_state+0x5c8)[0x6f9ad8] /usr/sbin/mysqld[0x6f9bdd] /usr/sbin/mysqld(_Z16dispatch_command19enum_server_commandP3THDPcj+0xc4e)[0x6facbe] /usr/sbin/mysqld(_Z10do_commandP3THD+0x159)[0x6fbc89] /usr/sbin/mysqld(_Z24do_handle_one_connectionP3THD+0x27d)[0x6cdaed] /usr/sbin/mysqld(handle_one_connection+0x42)[0x6cdb72] /usr/sbin/mysqld(pfs_spawn_thread+0x140)[0xb0f410] /lib/x86_64-linux-gnu/libpthread.so.0(+0x6b50)[0x7fe524b7cb50] /lib/x86_64-linux-gnu/libc.so.6(clone+0x6d)[0x7fe522e6920d] Trying to get some variables. Some pointers may be invalid and cause the dump to abort. Query (565aa00): START TRANSACTION /*!40100 WITH CONSISTENT SNAPSHOT */ Connection ID (thread ID): 6965 Status: NOT_KILLED You may download the Percona XtraDB Cluster operations manual by visiting http://www.percona.com/software/percona-xtradb-cluster/. You may find information in the manual which will help you identify the cause of the crash. 140826 11:42:49 mysqld_safe Number of processes running now: 0 140826 11:42:49 mysqld_safe WSREP: not restarting wsrep node automatically 140826 11:42:49 mysqld_safe mysqld from pid file /srv/databases/mysql/data/node3.pid ended
Since we’re currently running without backups, any quick help is highly appreciated. Please tell me if you need any more information.