KingbaseES V8R3 集群运维系列 -- db vip和cluster vip管理
案例说明：
KingbaseES V8R3集群集成了DB VIP（应用连接）和Cluster VIP（集群管理），本案例描述了两种vip在集群的相关配置及集群故障时vip漂移的问题。
适用版本：
KingbaseES V8R3
集群架构：
一、集群VIP相关配置
1）HAmodule.conf配置DB VIP和Cluster VIP
[kingbase@node101 bin]$ cat ../etc/HAmodule.conf |grep -i vip
#vip is bound to the specified network card.example:DEV="ens33"
#db use vip/the subnet mask.example:KB_VIP="192.168.28.220/24"
KB_VIP="192.168.1.204/24" #db vip配置
#pool use vip/the subnet mask.example:KB_POOL_VIP="192.168.28.220/24"
KB_POOL_VIP="192.168.1.205" #Cluster vip配置
---集群脚本kingbase_monitor.sh在执行时，会读取HAmodule.conf中配置信息。
2）kingbasecluster.conf中Cluster vip配置
[kingbase@node101 etc]$ cat kingbasecluster.conf|grep -i 'ip add'|grep -v '#'
if_up_cmd='ip addr add 192.168.1.205/24 dev enp0s3 label enp0s3:0'
if_down_cmd='ip addr del 192.168.1.205/24 dev enp0s3'
---在执行kingbasecluster启动或停止集群服务时，会读取kingbasecluster.conf中的配置，加载或卸载Cluster vip。
二、集群VIP加载
1）DB VIP加载
[kingbase@node101 bin]$ ./kingbase_monitor.sh start
-----------------------------------------------------------------------
2023-02-14 19:00:25 KingbaseES automation beging...
......................
ADD VIP NOW AT 2023-02-14 19:00:33 ON enp0s3
execute: [/sbin/ip addr add 192.168.1.204/24 dev enp0s3 label enp0s3:2]
execute: /home/kingbase/cluster/HAR3/db/bin//arping -U 192.168.1.204 -I enp0s3 -w 1
.....
all started..
---如上所示，执行kingbase_monitor.sh start时，DB vip被加载到集群数据库服务的主节点（Primary）上。
2）Cluster vip加载（cluster.log）
2023-02-14 19:01:00: pid 31342: LOG: kingbasecluster successfully started. version 3.6.7 (release)
.......
2023-02-14 19:01:02: pid 31449: LOG: successfully acquired the delegate IP:"192.168.1.205"
2023-02-14 19:01:02: pid 31449: DETAIL: 'if_up_cmd' returned with success
---如上，在cluster.log中显示，在kingbaseclsuter服务启动时，将读取kingbasecluster.conf配置，加载Cluster vip到集群主节点上.
3）查看主节点ip信息
[kingbase@node101 bin]$ ip add sh
......
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 08:00:27:bd:83:57 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.101/24 brd 192.168.1.255 scope global noprefixroute enp0s3
valid_lft forever preferred_lft forever
inet 192.168.1.204/24 scope global secondary enp0s3:2
valid_lft forever preferred_lft forever
inet 192.168.1.205/24 scope global secondary enp0s3:0
valid_lft forever preferred_lft forever
---如上所示，在主节点DB VIP和Cluster vip都已经被加载。
三、VIP地址漂移测试
1、Cluster vip漂移
1）模拟kingbasecluster服务down
# 查看kingbasecluster进程
[kingbase@node101 bin]$ ps -ef |grep kingbase
.......
root 31342 1 0 19:00 ? 00:00:00 ./kingbasecluster -n
root 31383 31342 0 19:00 ? 00:00:00 kingbasecluster: watchdog
root 31450 31342 0 19:01 ? 00:00:00 kingbasecluster: lifecheck
root 31452 31450 0 19:01 ? 00:00:00 kingbasecluster: heartbeat receiver
root 31453 31450 0 19:01 ? 00:00:00 kingbasecluster: heartbeat sender
root 31456 31342 0 19:01 ? 00:00:00 kingbasecluster: wait for connection request
root 31457 31342 0 19:01 ? 00:00:00 kingbasecluster: wait for connection request
.......
# kill kingbasecluster进程
[root@node101 ~]# kill -2 31342
2）查看集群节点vip信息
#原主节点
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 08:00:27:bd:83:57 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.101/24 brd 192.168.1.255 scope global noprefixroute enp0s3
valid_lft forever preferred_lft forever
inet 192.168.1.204/24 scope global secondary enp0s3:2
valid_lft forever preferred_lft forever
#原备节点
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 08:00:27:73:47:f6 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.102/24 brd 192.168.1.255 scope global noprefixroute enp0s3
valid_lft forever preferred_lft forever
inet 192.168.1.205/24 scope global secondary enp0s3:0
valid_lft forever preferred_lft forever
---如上所示，在主机点Cluster vip（192.168.1.205）已经被卸载，被加载到原备节点；
但是DB vip没有发生漂移，不影响应用对数据库服务的访问。
3）查看cluster.log日志
Tips：
如下所示，由于主节点kingbasecluster服务被停止，备库kingbasecluster服务切换为新主节点，Cluster vip漂移到了新的kingbasecluster主节点。
2023-02-14 19:02:51: pid 7497: LOG: We have lost the cluster master node "192.168.1.101:9999 Linux node101"
2023-02-14 19:02:51: pid 7497: LOG: watchdog node state changed from [STANDBY] to [JOINING]
......
2023-02-14 19:02:56: pid 7497: LOG: watchdog node state changed from [INITIALIZING] to [MASTER]
2023-02-14 19:02:56: pid 7497: LOG: I am announcing my self as master/coordinator watchdog node
2023-02-14 19:02:59: pid 7500: LOG: watchdog checking if kingbasecluster is alive using heartbeat
2023-02-14 19:02:59: pid 7500: DETAIL: the last heartbeat from "192.168.1.101:9999" received 8 seconds ago
2023-02-14 19:03:00: pid 7497: LOG: I am the cluster leader node
2023-02-14 19:03:00: pid 7497: DETAIL: our declare coordinator message is accepted by all nodes
........
2023-02-14 19:03:02: pid 8176: LOG: selecting backend connection
2023-02-14 19:03:02: pid 8176: DETAIL: failback event detected, discarding existing connections
2023-02-14 19:03:02: pid 7500: LOG: watchdog checking if kingbasecluster is alive using heartbeat
2023-02-14 19:03:02: pid 7500: DETAIL: the last heartbeat from "192.168.1.101:9999" received 11 seconds ago
2023-02-14 19:03:02: pid 9330: LOG: successfully acquired the delegate IP:"192.168.1.205"
2023-02-14 19:03:02: pid 9330: DETAIL: 'if_up_cmd' returned with success
4）重启原主节点的kingbasecluster服务
#启动kingbasecluster服务
[root@node101 ~]# cd /home/kingbase/cluster/HAR3/kingbasecluster/bin
[root@node101 bin]# ./restartcluster.sh
Tips：
如下所示，原主节点在启动kingbasecluster服务后，做为standby节点加入集群。
#cluster.log:
2023-02-14 19:03:05: pid 1023: LOG: watchdog node state changed from [DEAD] to [LOADING]
2023-02-14 19:03:05: pid 1023: LOG: new outbound connection to 192.168.1.102:9000
2023-02-14 19:03:05: pid 1023: LOG: setting the remote node "192.168.1.102:9999 Linux node102" as watchdog cluster master
2023-02-14 19:03:05: pid 1023: LOG: watchdog node state changed from [LOADING] to [INITIALIZING]
2023-02-14 19:03:05: pid 1023: LOG: new watchdog node connection is received from "192.168.1.102:47600"
2023-02-14 19:03:05: pid 1023: LOG: new node joined the cluster hostname:"192.168.1.102" port:9000 kingbasecluster_port:9999
2023-02-14 19:03:06: pid 1023: LOG: watchdog node state changed from [INITIALIZING] to [STANDBY]
2023-02-14 19:03:06: pid 1023: LOG: successfully joined the watchdog cluster as standby node
2、DB VIP漂移
1）模拟主库数据库服务down
[kingbase@node101 bin]$ ./sys_ctl stop -D /home/kingbase/cluster/HAR3/db/data
2）查看failover.log日志
-----------------2023-02-14 19:23:52 failover beging---------------------------------------
----failover-stats is %H = hostname of the new master node [192.168.1.102], %P = old primary node id [0], %d = node id[0], %h = host name [192.168.1.101], %O = old primary host[192.168.1.101] %m = new master node id [1], %M = old master node id [0], %D = database cluster path [/home/kingbase/cluster/HAR3/db/data].
----ping trust ip
ping trust ip 192.168.1.1 success ping times :[3], success times:[3]
----determine whether the faulty db is master or standby
master down, let 192.168.1.102 become new primary.....
2023-02-14 19:23:54 del old primary VIP on 192.168.1.101
es_client connect host:192.168.1.101 success, will stop old primary db and del the vip
stop the old primary db
sys_ctl: PID file "/home/kingbase/cluster/HAR3/db/data/kingbase.pid" does not exist
Is server running?
DEL VIP NOW AT 2023-02-14 19:23:56 ON enp0s3
execute: [/sbin/ip addr del 192.168.1.204/24 dev enp0s3]
Oprate del ip cmd end.
2023-02-14 19:23:54 add VIP on 192.168.1.102
ADD VIP NOW AT 2023-02-14 19:23:55 ON enp0s3
execute: [/sbin/ip addr add 192.168.1.204/24 dev enp0s3 label enp0s3:2]
execute: /home/kingbase/cluster/HAR3/db/bin//arping -U 192.168.1.204 -I enp0s3 -w 1
Success to send 1 packets
2023-02-14 19:23:55 promote begin...let 192.168.1.102 become master
.......
-----------------2023-02-14 19:23:55 failover end---------------------------------------
---如上所示，failover切换过程中，DB VIP将从原主库卸载，新主库加载。
3）查看新主库ip信息
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 08:00:27:73:47:f6 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.102/24 brd 192.168.1.255 scope global noprefixroute enp0s3
valid_lft forever preferred_lft forever
inet 192.168.1.205/24 scope global secondary enp0s3:0
valid_lft forever preferred_lft forever
inet 192.168.1.204/24 scope global secondary enp0s3:2
---如上所示，在集群触发failover切换后，DB VIP漂移到新的primary节点。
四、总结
KingbaseES V8R3集群通过vip地址，实现应用对数据库高可用性的连接访问及集群管理。
1）DB VIP用于应用的连接访问，在启动集群时被加载到数据库服务的主节点（Primary），当主节点数据库服务down机，触发failover切换时，DB VIP漂移到新的数据库服务主节点。
2）Cluster vip用于kingbasecluster服务的访问，集群启动时加载到kingbaseclsuter的Master节点，当master节点的kingbasecluster服务down时，会漂移到新的master节点。
3）在生产环境出现不能访问9999端口（kingbasecluster服务端口）时，可以尝试重启kingbasecluster服务，默认是不影响客户端的连接访问数据库服务；但对于生产环境，最好是在应用访问的低峰时间执行。
KINGBASE研究院
