MyCat 学习笔记 第十九篇 . mycat 监控工具 mycat web
mycat 自带了一个简单的监控工具 mycat eye（自带了 jetty 的一个web 应用）。可以支持 mycat 环境参数、mycat 运行状态、 mysql 同步状态、高瓶 sql 等数据的实时查询。相对于mycat server来说，mycat web 发展的非常缓慢，目前可以定义为是mycat运行情况的辅助查询工具，可以替代少数 9066 管理端口做的事务。
mycat web 使用了 zookeeper 命名服务功能，把相关的mycat 配置信息存放到 zk 一部分，另一部分存放在mycat web本地的sql lite中，估计是当初设计不充分~~~
mycat web 下载地址
github 入口
另外有四个相对重要的文件
| 文件 | 用处 |
|---|---|
| mycat_home\bin\xml_to_yaml.sh | 根据mycat配置生成 zookeeper yaml 配置工具 |
| mycat_home\bin\init_zk_data.sh | 将生成的yaml配置文件导入远程zookeeper |
| mycat_home\conf\myid.properties | zookeeper 路径配置参数信息 |
| mycat_home\conf\zk-create.yaml | 自动生成的zk配置 |
安装顺序
1）确认安装好zookeeper 环境（此步略，网上教程一大把）。
2）启动mycat服务，确认本地服务一切正常。
3）执行xml_to_yaml.sh脚本（确认本地配置过 JAVA_HOME 和 MYCAT_HOME环境变量，没有的话参考下文）
/etc/profile配置
export JAVA_HOME=/usr/java/jdk1.7.0_79/
export CLASSPATH=.:%JAVA_HOME%/lib/dt.jar:%JAVA_HOME%/lib/tools.jar
export PATH=$PATH:$JAVA_HOME/bin
export MYCAT_HOME=/usr/local/mycat/mycat-server-1.5/4）查看yaml配置是否正确（基本看不出问题，因为配置有点重杂了~~）
zkURL: 127.0.0.1:2181
mycat-mysqlgroup:
host_183:
name: host_183
repType: '0'
zone: fz
servers:
- host_183
cur-write-server: host_183
auto-write-switch: true
heartbeatSQL: select user()
host_185:
name: host_185
repType: '0'
zone: fz
servers:
- host_185
cur-write-server: host_185
auto-write-switch: true
heartbeatSQL: select user()
mycat-mysqls:
host_183:
name: host_183
ip: 192.168.13.191
port: '3307'
user: root
password: root123
hostId: host
zone: fz
host_185:
name: host_185
ip: 192.168.13.185
port: '3307'
user: root
password: root123
hostId: host
zone: fz
mycat-cluster:
mycat-cluster-1:
user:
test:
name: test
password: test
schemas:
- TESTDB
user:
name: user
password: user
schemas:
- TESTDB
readOnly: 'true'
rule:
sharding-by-date:
name: partbymonth
functionName: org.opencloudb.route.function.PartitionByMonth
column: date_str
dateFormat: yyyy-MM-dd
sBeginDate: '2015-01-01'
rule1:
name: func1
functionName: org.opencloudb.route.function.PartitionByLong
column: id
partitionCount: '8'
partitionLength: '128'
rule2:
name: func1
functionName: org.opencloudb.route.function.PartitionByLong
column: user_id
partitionCount: '8'
partitionLength: '128'
sharding-by-intfile:
name: hash-int
functionName: org.opencloudb.route.function.PartitionByFileMap
column: sharding_id
config:
'10000': '0'
'10010': '1'
auto-sharding-long:
name: rang-long
functionName: org.opencloudb.route.function.AutoPartitionByLong
column: id
config:
0-500M: '0'
500M-1000M: '1'
1000M-1500M: '2'
mod-long-userId:
name: mod-long
functionName: org.opencloudb.route.function.PartitionByMod
column: user_id
count: '3'
mod-long:
name: mod-long
functionName: org.opencloudb.route.function.PartitionByMod
column: id
count: '3'
sharding-by-murmur:
name: murmur
functionName: org.opencloudb.route.function.PartitionByMurmurHash
column: id
seed: '0'
count: '2'
virtualBucketTimes: '160'
sharding-by-month:
name: partbymonth
functionName: org.opencloudb.route.function.PartitionByMonth
column: create_date
dateFormat: yyyy-MM-dd
sBeginDate: '2015-01-01'
latest-month-calldate:
name: latestMonth
functionName: org.opencloudb.route.function.LatestMonthPartion
column: calldate
splitOneDay: '24'
auto-sharding-rang-mod:
name: rang-mod
functionName: org.opencloudb.route.function.PartitionByRangeMod
column: id
config:
0-200M: '5'
200M1-400M: '1'
400M1-600M: '4'
600M1-800M: '4'
800M1-1000M: '6'
jch:
name: jump-consistent-hash
functionName: org.opencloudb.route.function.PartitionByJumpConsistentHash
column: id
totalBuckets: '3'
mod-sharding-long:
name: mod-long
functionName: org.opencloudb.route.function.PartitionByMod
column: sharding_mod_id
count: '3'
schema:
TESTDB:
name: TESTDB
checkSQLSchema: true
defaultMaxLimit: 100
new_table:
name: new_table
datanode: dn1,dn2,dn3,dn4,dn5,dn6,dn7,dn8,dn9,dn10,dn11,dn12
ruleName: sharding-by-date
t_user:
name: t_user
datanode: dn1,dn2,dn3
ruleName: mod-long-userId
t_order:
name: t_order
datanode: dn1,dn2,dn3
ruleName: mod-long-userId
t_order_fee:
name: t_order_fee
datanode: dn1,dn2,dn3
ruleName: mod-long-userId
datanode:
dn1:
name: dn1
database: ds1
dataHost: '13_183'
dn2:
name: dn2
database: ds2
dataHost: '13_183'
dn3:
name: dn3
database: ds3
dataHost: '13_183'
dn4:
name: dn4
database: ds4
dataHost: '13_183'
dn5:
name: dn5
database: ds5
dataHost: '13_183'
dn6:
name: dn6
database: ds6
dataHost: '13_183'
dn7:
name: dn7
database: ds7
dataHost: '13_183'
dn8:
name: dn8
database: ds8
dataHost: '13_183'
dn9:
name: dn9
database: ds9
dataHost: '13_183'
dn10:
name: dn10
database: ds10
dataHost: '13_183'
dn11:
name: dn11
database: ds11
dataHost: '13_185'
dn12:
name: dn12
database: ds12
dataHost: '13_185'
datahost:
'13_183':
name: '13_183'
balance: 0
maxcon: 1000
mincon: 10
dbtype: mysql
dbDriver: native
writeType: 0
switchType: 1
slaveThreshold: 100
heartbeatSQL: select user()
mysqlGroup: host_183
'13_185':
name: '13_185'
balance: 0
maxcon: 1000
mincon: 10
dbtype: mysql
dbDriver: native
writeType: 0
switchType: 1
slaveThreshold: 100
heartbeatSQL: select user()
mysqlGroup: host_185
mycat-nodes:
mycat_fz_01:
name: mycat_fz_01
cluster: mycat-cluster-1
systemParams:
defaultSqlParser: druidparser
5 ) 执行 init_zk_data.sh 脚本 , 注意下默认情况下脚本会将zookeeper定位在同一台服务器上面，如有需要可以调整下 ZK 的 IP : PORT
demo.catlets.ZkCreate
public static void main(String[] args) {
String url;
if (args != null && args.length > 0) {
ZK_CONFIG_FILE_NAME = args[0];
url = args[1];
} else {
url = zkConfig.containsKey(CONFIG_URL_KEY) ?
(String) zkConfig.get(CONFIG_URL_KEY) :
"127.0.0.1:2181"; ----- 这里居然写死掉了 ^=^
}
zkConfig = loadZkConfig();
framework = createConnection(url);
createConfig(MYCAT_HOST_KEY, false, MYCAT_HOST_KEY);
createConfig(MYCAT_ZONE_KEY, false, MYCAT_ZONE_KEY);
createConfig(MYCAT_NODES_KEY, false, MYCAT_NODES_KEY);
createConfig(MYCAT_CLUSTER_KEY, true, MYCAT_CLUSTER_KEY);
createConfig(MYCAT_MYSQLS_KEY, true, MYCAT_MYSQLS_KEY);
createConfig(MYCAT_MYSQL_GROUP_KEY, true, MYCAT_MYSQL_GROUP_KEY);
createConfig(MYCAT_LBS, true, MYCAT_LBS);
}
6 ) 配置 mycat-web 中 zk 的地址
/usr/local/mycat/mycat-web/mycat-web/WEB-INF/classes
#
#Mon Jan 16 15:37:36 CST 2012
show.period=3000000
zookeeper=192.168.13.185:2181
##sql\u4E0A\u7EBF\u76F8\u5173\u914D\u7F6E
sqlonline.server=192.168.13.187
sqlonline.user=root
sqlonline.passwd=123456
7 ) 启动 mycat web ，这里为了方便查询日志就直接通脚本启动
java -Xms1024m -Xmx1024m -Xmn128m -XX:PermSize=64m -XX:+UseParNewGC -XX:+UseConcMarkSweepGC -XX:+UseCMSCompactAtFullCollection -XX:CMSFullGCsBeforeCompaction=0 -XX:CMSInitiatingOccupancyFraction=70 -Dcom.sun.management.jmxremote -Dcom.sun.management.jmxremote.port=8999 -Dcom.sun.management.jmxremote.authenticate=false -Dcom.sun.management.jmxremote.ssl=false -jar start.jar
部分功能效果图
为了看下mycat web的效果，压了100万条记录，可以比较真观的看到所有SQL的执行情况和较慢的SQL的运行时长，不能算是太鸡肋吧，但要全面运维一个服务来说，mycat web 还有比较长的路需要走。
OK，今天就先到这里~~