Web lists-archives.com

Re: can't start mariadb with client certs specified

Hi Reindl,

what about running mysql_upgrade *directly* after the major update and
> *before* touch anything else?

That was precisely what happened. In setting up the new database machine,
puppet had installed version 5 of mariadb. Before even starting up version
of for the first time, I uninstalled the RPM's for that version and
installed version 10.0.20 from the mariadb repo.

However I was able to solve this problem my recreating the keys and certs.
I had setup master/slave replication using this tutorial:


I'm not sure what the problem is with the way the keys are generated in
this article that could have caused that last, rather long email. :)

So after I regenerated the keys and certs using this method:

  182  openssl genrsa -des3 -out db1.example.com.key 4096
  183  openssl rsa -in db1.example.com.key -out db1.example.com.key
  184  openssl req -new -key db1.example.com.key -out db1.jokefire.com.csr
  185  openssl x509 -req -days 3650 -in db1.example.com.csr -CA ca.crt
-CAkey ca.key -set_serial 01 -out db1.example.com.crt

I was able to start up both db's 1 and 2 with the SSL keys and certs
specified in both the client and server sections of the my.cnf. Like this
example from the first node shows:

[root@db1:~] #egrep -v '^#|^$'  /etc/my.cnf

















Anyplace you see a line of dots there was omitted to save space.

I am facing a new issue however. It looks as if I've been able to get slave
replication working as I try to demonstrate below:

[root@db2:~] #mysql -e "show slave status \G" | egrep -i

               Slave_IO_State: Waiting for master to send event

              Master_Log_File: mariadb-bin.000004

        Relay_Master_Log_File: mariadb-bin.000004

             Slave_IO_Running: Yes

            Slave_SQL_Running: Yes

        Seconds_Behind_Master: 0

But I'm seeing an error that I've never seen before relating to 'gtid':

[root@db2:~] #mysql -e "show slave status \G" | egrep -i "error|errno"

                   Last_Errno: 1146

                   *Last_Error: Unable to load replication GTID slave state
from mysql.gtid_slave_pos: Table 'mysql.gtid_slave_pos' doesn't exist*

                Last_IO_Errno: 0


               Last_SQL_Errno: 1146

              * Last_SQL_Error: Unable to load replication GTID slave state
from mysql.gtid_slave_pos: Table 'mysql.gtid_slave_pos' doesn't exist*

Any idea what that error means and how I can get rid of it? I've done some
reading on the error and tried setting my gtid variables for both nodes.
For example this is what I have in the first node:

MariaDB [(none)]> show variables like '%gtid%';


| Variable_name          | Value |


| gtid_binlog_pos        |       |

| gtid_binlog_state      |       |

| gtid_current_pos       |       |

| gtid_domain_id         | 1     |

| gtid_ignore_duplicates | OFF   |

| gtid_seq_no            | 0     |

| gtid_slave_pos         |       |

| gtid_strict_mode       | OFF   |

| last_gtid              |       |


9 rows in set (0.00 sec)

Can you please help me out with this one as well?



On Wed, Jul 22, 2015 at 4:31 AM, Reindl Harald <h.reindl@xxxxxxxxxxxxx>

> Am 22.07.2015 um 06:39 schrieb Tim Dunphy:
>> Hey guys,
>> I'm trying to setup multi master replication using SSL under MariaDB
>> 10.20.
>> I've been able to specify my client certs on the second node (db2) with no
>> issue and start up the mysql service
> what about running mysql_upgrade *directly* after the major update and
> *before* touch anything else?
> 150722  4:18:47 [ERROR] Column count of mysql.file_summary_by_event_name is
> wrong. Expected 23, found 5. Created with MariaDB 50541, now running
> 100020. Please use mysql_upgrade to fix this error

GPG me!!

gpg --keyserver pool.sks-keyservers.net --recv-keys F186197B