DRBD Basic (Implement database high availability)

DRBD (Implement database high availability)

Distributed replicated block Device (DRBD) is a software-implemented, non-shared storage replication solution for mirrored block device content between servers.

Data mirroring: Real-time, transparent, synchronous (all servers are returned after success), asynchronous (returned after successful local server)

Environment

node1:192.168.163.172

node2:192.168.163.173

installation process

Modify the Hosts file first

[Email protected] ~]# cat/etc/hosts

127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4

:: 1 localhost localhost.localdomain localhost6 localhost6.localdomain6

192.168.163.172 Node1

192.168.163.173 Node2

[Email protected] ~]# cat/etc/hosts

127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4

:: 1 localhost localhost.localdomain localhost6 localhost6.localdomain6

192.168.163.172 Node1

192.168.163.173 Node2

Host name also to be modified

[Email protected] ~]# hostname

Node1

[Email protected] ~]# hostname

Node2

SSH-free login on both sides

[Email protected] ~]# Ssh-copy-id Node2

[Email protected] ~]# Ssh-copy-id Node1

To set the clock synchronization:

[Email protected] ~]# crontab-l

*/5 * * * * ntpdate cn.pool.ntp.org

[Email protected] ~]# crontab-l

*/5 * * * * ntpdate cn.pool.ntp.org

Install Derb:

# RPM--import https://www.elrepo.org/RPM-GPG-KEY-elrepo.org

# RPM-UVH http://www.elrepo.org/elrepo-release-7.0-2.el7.elrepo.noarch.rpm

# yum Install-y kmod-drbd84 drbd84-utils

To edit a global configuration file

Vim/etc/drbd.d/global_common.conf

global {    usage-count no;   #是否参加DRBD使用统计, the default is yes. Official statistics DRBD Installed Capacity     # minor-count dialog-refresh disable-ip-verification}common  {    protocol C;       #使用DRBD的同步协议      handlers {        pri-on-incon-degr  "/usr/lib/ Drbd/notify-pri-on-incon-degr.sh; /usr/lib/drbd/notify-emergency-reboot.sh; echo b >  /proc/sysrq-trigger ; reboot -f ";         pri-lost-after-sb  "/usr/lib/drbd/notify-pri-lost-after-sb.sh; /usr/lib/drbd/ Notify-emergency-reboot.sh; echo b > /proc/sysrq-trigger ; reboot -f ";         local-io-error  "/usr/lib/drbd/notify-io-error.sh; / Usr/lib/drbd/notify-emergency-shutdown.sh; echo o >&nbsP;/proc/sysrq-trigger ; halt -f ";    }    startup  {        # wfc-timeout degr-wfc-timeout  outdated-wfc-timeout wait-after-sb    }    options {         # cpu-mask on-no-data-accessible    }     disk {        on-io-error detach;   #配置I/O error handling policy for detach         # size max-bio-bvecs  on-io-error fencing disk-barrier disk-flushes        #  disk-drain md-flushes resync-rate resync-after al-extents         # c-plan-ahead c-delay-target c-fill-target c-max-rate         # c-min-rate disk-timeout    }    net {         # protocol timeout max-epoch-size max-buffers  unplug-watermark        # connect-int ping-int  sndbuf-size rcvbuf-size ko-count        #  allow-two-primaries cram-hmac-alg shared-secret after-sb-0pri         # after-sb-1pri after-sb-2pri always-asbp rr-conflict         # ping-timeout data-integrity-alg tcp-cork on-congestion         # congestion-fill congestion-extents csums-alg  verify-alg        # use-rle    }     syncer {        rate 1024m;     #设置主备节点同步时的网络速率     }} 

Creating a configuration file

[Email protected] ~]# Vim/etc/drbd.d/mysql.res

Resource MySQL {protocol c;meta-disk internal;device/dev/drbd1;syncer {verify-alg sha1;} NET {allow-two-primaries;} On node1 {disk/dev/sdb2; #在node1创建的分区address 192.168.163.172:7789;} On node2 {disk/dev/sdb2; #在node2创建的分区address 192.168.163.173:7789;}}

Then copy the Node1 configuration file to the Node2:

scp-rp/etc/drbd.d/* node2:/etc/drbd.d/

Start on Node2:

[[email protected] ~]# drbdadm CREATE-MD MySQL

Initializing activity log

Initializing bitmap (KB) to all zero

Writing Meta Data ...

New DRBD Meta data block successfully created.

[Email protected] ~]# modprobe DRBD

To see if the kernel has loaded the module:

[Email protected] drbd.d]# Lsmod | grep DRBD

DRBD 396875 1

LIBCRC32C 12644 4 Xfs,drbd,ip_vs,nf_conntrack

###

[[email protected] ~]# drbdadm up MySQL

[[email protected] ~]# DRBDADM----Force primary MySQL

View Status:

[Email protected] ~]# CAT/PROC/DRBD

Version:8.4.10-1 (api:1/proto:86-101)

GIT-HASH:A4D5DE01FFFD7E4CDE48A080E2C686F9E8CEBF4C build by [email protected], 2017-09-15 14:23:22

1:cs:wfconnection Ro:primary/unknown Ds:uptodate/dunknown C r-----

ns:0 nr:0 dw:0 dr:912 al:0 bm:0 lo:0 pe:0 ua:0 ap:0 ep:1 wo:f oos:0

The same operation is performed again in Node2

[[email protected] ~]# drbdadm CREATE-MD MySQL

[Email protected] ~]# modprobe DRBD

[[email protected] ~]# drbdadm up MySQL

Both sides are installed MARIADB

Yum install-y mariadb Mariadb-server

Modify the configuration file to point the database path to the/test directory

Vim/etc/my.cnf

[Mysqld]datadir=/testsocket=/var/lib/mysql/mysql.sock

View status, see Node1 is primary

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon. Please consider using Drbdtop. 1:mysql/0 Connected primary/secondary uptodate/uptodate

Create a/test directory

Mkdir/test

format and Mount in Node1 :

[Email protected] ~]# MKFS.XFS/DEV/DRBD1

[Email protected] ~]# mount/dev/drbd1/test

Mount succeeded

[Email protected] ~]# df-h

Filesystem Size used Avail use% mounted on

/dev/mapper/cl-root 17G 1.5G 16G 9%/

Devtmpfs 902M 0 902M 0%/dev

Tmpfs 912M 0 912M 0%/dev/shm

Tmpfs 912M 8.5M 904M 1%/run

Tmpfs 912M 0 912M 0%/sys/fs/cgroup

/DEV/SDA1 1014M 185M 830M 19%/boot

Tmpfs 183M 0 183M 0%/run/user/0

/DEV/DRBD1 991M 32M 892M 4%/test

=================================================

Test session:

Start the database first and create a database

[Email protected] ~]# systemctl start mariadb

MariaDB [(None)]> CREATE Database xhk777;

Query OK, 1 row affected (0.04 sec)

Stop the Node1 mariadb service and uninstall the mount you just mounted

[Email protected] ~]# Systemctl stop mariadb

[Email protected] ~]# umount/test

Downgrade Node1 to Secondary

[[email protected] ~]# drbdadm secondary MySQL

Upgrade Node2 to Primary

[[email protected] ~]# Drbdadm primary MySQL

Note: In single-master mode, DRBD, two nodes are connected at the same time, any one node can become the master in a certain time, but only one of the two nodes, if there is already a master, need to downgrade before possible upgrade; there is no limit in dual-master mode

Check the status on Node2 and you can see that Node2 is already primary

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 Connected primary/secondary uptodate/uptodate

Mount at this time and start the database

[Email protected] ~]# mount/dev/drbd1/test

[Email protected] ~]# systemctl start mariadb

View the database and find the xhk777 you just created

MariaDB [(None)]> show databases;+---------------------+| Database |+---------------------+| Information_schema | | #mysql50 #lost+found | | MySQL | | Performance_schema | | XHK | | xhk777 |+---------------------+6 rows in Set (0.06 sec)

The experiment was successful, and we could simulate a brain fissure after the node1 recovered.

Prohibit communication above the Node2

[Email protected] ~]# iptables-p INPUT DROP

When Node1 view the state, it is not possible to find Node2.

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 wfconnection Secondary/Unknown Uptodate/dunknown

Upgrade Node1 to Primary

[[email protected] ~]# Drbdadm primary MySQL

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 wfconnection Primary/Unknown Uptodate/dunknown

Mount the directory as before and start the MARIADB service

[Email protected] ~]# mount/dev/drbd1/test

[Email protected] ~]# systemctl start mariadb

Node1 database service is OK

MariaDB [(None)]> show databases;+---------------------+| Database |+---------------------+| Information_schema | | #mysql50 #lost+found | | MySQL | | Performance_schema | | XHK | | xhk777 |+---------------------+6 rows in Set (0.05 sec)

This time to change the Node2 firewall policy to allow, there will be a brain fissure situation

Iptables-p INPUT ACCEPT

View the status separately and discover that Node1 and Node2 are all for primary

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 StandAlone Primary/Unknown uptodate/dunknown/test ext4 991M 32M 892M 4%

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 StandAlone Primary/Unknown uptodate/dunknown/test ext4 991M 32M 892M 4%

The workaround is to use the Node1 as a backup node

[Email protected] ~]# Systemctl stop mariadb

[Email protected] ~]# umount/test

[[email protected] ~]# drbdadm secondary MySQL

[[email protected] ~]# drbdadm connect--discard-my-data mysql

View the status of Node1

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 wfconnection Secondary/Unknown Uptodate/dunknown

Discover or unknown, reconnect to Node2

[[email protected] ~]# drbdadm connect MySQL

View status again

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 Connected Secondary/Primary uptodate/uptodate

[Email protected] ~]# Drbd-overview

Note:drbd-overview'll be deprecated soon.

Please consider using Drbdtop.

1:mysql/0 Connected Primary/secondary uptodate/uptodate/test ext4 991M 32M 892M 4%

Back to normal!!!

This article is from the "xhk777" blog, make sure to keep this source http://xhk777.blog.51cto.com/13405744/1975381

DRBD Basic (Implement database high availability)