[MXS-314] Read Write Split Error with Galera Nodes Created: 2015-08-17  Updated: 2015-09-17  Resolved: 2015-08-21

Status: Closed
Project: MariaDB MaxScale
Component/s: readwritesplit
Affects Version/s: 1.2.0
Fix Version/s: 1.2.1

Type: Bug Priority: Blocker
Reporter: christian stehle Assignee: Johan Wikman
Resolution: Fixed Votes: 0
Labels: None
Environment:

Centos 7, 3x Galera Nodes with MariaDB Galera 10.0.20-1


Attachments: Text File debug1.log     Text File trace1.log    
Issue Links:
Duplicate
duplicates MXS-270 Crash with MySQLBackend protocol module Closed

 Description   

With a Typo3 System using Read Write Splitting I regulary get erros and maxscale-restarts:

Edit: I noticed that I also get the errors with Master/slave Configuration.

Error-log:

{{MariaDB Corporation MaxScale /var/log/maxscale/error1.log Mon Aug 17 11:16:20 2015
-----------------------------------------------------------------------
— Logging is enabled.
2015-08-17 11:16:20 Error: Not a boolean value:
2015-08-17 11:16:20 Error: Failed to set socket options. Error 95: Operation not supported
2015-08-17 11:19:08 Fatal: MaxScale received fatal signal 11. Attempting backtrace.
2015-08-17 11:19:08 /usr/bin/maxscale() [0x549687]
2015-08-17 11:19:08 /lib64/libpthread.so.0(+0xf130) [0x7f9bfe093130]
2015-08-17 11:19:08 /usr/lib64/maxscale/libMySQLBackend.so(+0x6fbb) [0x7f9be04b9fbb]
2015-08-17 11:19:08 /usr/lib64/maxscale/libMySQLBackend.so(+0x4fef) [0x7f9be04b7fef]
2015-08-17 11:19:08 /usr/bin/maxscale() [0x55ecaa]
2015-08-17 11:19:08 /usr/bin/maxscale(poll_waitevents+0x616) [0x55e573]
2015-08-17 11:19:08 /lib64/libpthread.so.0(+0x7df5) [0x7f9bfe08bdf5]
2015-08-17 11:19:08 /lib64/libc.so.6(clone+0x6d) [0x7f9bfc95a1ad]}}

Server status:

{{Server 0x2616440 (server1)
Server: 192.168.xxx.xxx
Status: Slave, Synced, Running
Protocol: MySQLBackend
Port: 3206
Server Version: 10.0.20-MariaDB-wsrep-log
Node Id: 1
Master Id: -1
Repl Depth: 0
Number of connections: 0
Current no. of conns: 0
Current no. of operations: 0
Server 0x2616330 (server2)
Server: 192.168.xxx.xxx
Status: Slave, Synced, Running
Protocol: MySQLBackend
Port: 3206
Server Version: 10.0.20-MariaDB-wsrep-log
Node Id: 2
Master Id: -1
Repl Depth: 0
Number of connections: 19
Current no. of conns: 2
Current no. of operations: 0
Server 0x2572af0 (server3)
Server: 192.168.xxx.xxx
Status: Master, Synced, Running
Protocol: MySQLBackend
Port: 3206
Server Version: 10.0.20-MariaDB-wsrep-log
Node Id: 0
Master Id: -1
Repl Depth: 0
Number of connections: 19
Current no. of conns: 2
Current no. of operations: 0}}

Maxscale Config:

[maxscale]
threads=4

[Galera Monitor]
type=monitor
module=galeramon
servers=server1,server2,server3
user=maxscale
passwd=xxxx
monitor_interval=10000
disable_master_failback=

[qla]
type=filter
module=qlafilter
options=/tmp/QueryLog

[fetch]
type=filter
module=regexfilter
match=fetch
replace=select

[CLI]
type=service
router=cli
[CLI Listener]
type=listener
service=CLI
[CLI]
type=service
router=cli

[CLI Listener]
type=listener
service=CLI
protocol=maxscaled
port=6603

[RW Split Router]
type=service
router=readwritesplit
servers=server1,server2,server3
user=maxscale
passwd=xxx

[RW Split Listener]
type=listener
service=RW Split Router
protocol=MySQLClient
port=3310
socket=/tmp/ClusterMaster

[server1]
type=server
address=192.168.xxx.xxx
port=3206
protocol=MySQLBackend

[server2]
type=server
address=192.168.xxx.xxx
port=3206
protocol=MySQLBackend

[server3]
type=server
address=192.168.xxx.xxx
port=3206
protocol=MySQLBackend



 Comments   
Comment by Dipti Joshi (Inactive) [ 2015-08-17 ]

Assigning to johan.wikman, Massimiliano Pinto and tturenko please work with Johan to help him set this up and troubleshoot it.

Thanks,
Dipti

Comment by markus makela [ 2015-08-17 ]

The crash seems to have happened in mysql_backend.c when preparing a prepared statement and the result set from the server was not contained in one network packet. The packets were evenly split so that there were no leftover bytes from the first part of the transmission.

Comment by Dipti Joshi (Inactive) [ 2015-08-17 ]

markus makela Can you please reproduce this along with Johan and review the code changes that are needed ?

Comment by christian stehle [ 2015-08-18 ]

See debug and trace logs before crash

Comment by Dipti Joshi (Inactive) [ 2015-08-21 ]

markus makela is the fix you did for prepared statements takes care of this ?

Thanks,
Dipti

Comment by markus makela [ 2015-08-21 ]

Yes, it is fixed. I'll close the issue.

Generated at Thu Feb 08 03:58:22 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.