Ganglia+Nagios监控系统

.安装httpd和phpyum -y install httpd php.安装依赖包yum -y install apr-devel apr-util check-devel cairo-devel pango-devel libxml2-devel rpm-build glib2-devel dbus-devel freetype-devel fontconfig-devel gcc gcc-c++ expat-devel python-devel libXrender-develyum -y install libart_lgpl-devel pcre-devel libtool rsyncyum -y install rrdtool rrdtool-devel #时序数据库.安装confusewget http://savannah.nongnu.org/download/confuse/confuse-2.7.tar.gztar xf confuse-2.7.tar.gzcd confuse-2.7/./configure --prefix=/usr/local/confuse CFLAGS=-fPIC --disable-nls --libdir=/usr/local/confuse/lib64make && make install.安装gangliawget https://sourceforge.net/projects/ganglia/files/ganglia%20monitoring%20core/3.7.2/ganglia-3.7.2.tar.gztar xf ganglia-3.7..tar.gzcd ganglia-3.7././configure --prefix=/usr/local/ganglia --enable-gexec --with-python --enable-status --with-gmetad --with-libconfuse=/usr/local/confuse
#Ganglia在安装时默认不安装gmetad，所以需要在configure时指定：--with-gmetadmake && make installcp gmetad/gmetad.init /etc/init.d/gmetadcp gmond/gmond.init /etc/init.d/gmond/usr/local/ganglia/sbin/gmond -t >/usr/local/ganglia/etc/gmond.conf #生成gmond.conf配置文件ln -s /usr/local/ganglia/sbin/gmetad /usr/sbin/gmetadln -s /usr/local/ganglia/sbin/gmond /usr/sbin/gmond.安装gwebwget https://sourceforge.net/projects/ganglia/files/ganglia-web/3.7.2/ganglia-web-3.7.2.tar.gztar xf ganglia-web-3.7..tar.gz -C /var/www/html/mv /var/www/html/ganglia-web-3.7. /var/www/html/gangliacd /var/www/html/gangliased -i 's/www-data/apache/' Makefile #用户默认为www-data,修改为apachemake install.修改配置(). 修改gmetad启动脚本# vim /etc/init.d/gmetad[ -f /usr/local/ganglia/etc/gmetad.conf ] || exit #这里将配置文件改成现在的位置，不然启动没反应(). 修改gmond启动文件vim /etc/init.d/gmond[ -f /usr/local/ganglia/etc/gmond.conf ] || exit #这里将配置文件改成现在的位置，不然启动没反应(). 创建rrds目录# mkdir /var/lib/ganglia/rrds -p #ganglia默认指定的rrds路径# chown -R nobody:nobody /var/lib/ganglia/rrds(.) 创建pid文件目录# mkdir -p /usr/local/ganglia/var/run/.gweb 添加密码认证(). 生成账号密码htpasswd -c /etc/httpd/auth.basic adminganglia(). vim /etc/httpd/conf/httpd.conf #在配置文件末尾加上下面这些：Alias /ganglia "/var/www/html/ganglia"<Directory "/var/www/html/ganglia"> AuthType Basic Options None AllowOverride None Order allow,deny Allow from all AuthName "ganglia Access" AuthUserFile/etc/httpd/auth.basic Require valid-user</Directory>.启动服务# systemctl start gmond.service# systemctl start gmetad.service# systemctl start httpd.service

(). 安装epel源wget -O /etc/yum.repos.d/epel.repo http://mirrors.aliyun.com/repo/epel-7.repoyum clean allyum makecache(). 下载并启动yum install -y ganglia-gmond ganglia-gmond-python #默认不支持自定义python监控，所以要添加。systemctl start gmond.service

gmetad.conf 配置文件是由单行属性及相应值组成。属性名不区分大小写，但属性值区分大小写。

#vim /usr/local/ganglia/etc/gmetad.conf.data_source属性 每一行data_source描述一个gmetad收集信息的gmond集群或gmetad网格。当gmetad检测到data_source引用一个集群时，gmetad将为data_source保留一整套轮询数据库（Round Robin Database）。然而，当gmetad检测一个data_source引用一个网格时，gmetad将只保留概要性的RRD。设置scalable属性为off时，gmetad 将强制为网格data_source保留一整套RRD文件，该值默认为on。下面是默认配置中的data_source配置示例：# data_source "my cluster" localhost my.machine.edu: 1.2.3.5:# data_source "my grid" 1.3.4.7: grid.org: grid-backup.org:# data_source "another source" 1.3.4.7: 1.3.4.8每个data_source有3个字段组成。第一个字段为唯一标识data_source的字符串；第二个字段为指定轮询间隔（单位是秒，默认为15秒，不应大于80秒）；第三个字段是以空格分开的所要轮询数据的主机列表，可以以IP地址或主机名的形式指定，同时可以指定端口号。如果未指定端口号，默认连接tcp/。.gridname属性 能唯一标识网格的字符串。该字符串应该与网络中其他gmetad所设置的标识符不同。.authority（URL） 网格的授权URL，被其他gmetad用来找到当前gmetad数据源的图表位置。默认值为“http://hostname/ganglia/”.trusted_hosts（文本格式） 允许其他gmetad获取本机8152端口的xml访问权限的主机列表，以空格作为分隔。Localhost总是可信的。.all_trusted（on|off） 当该值设置为on时，将重写trusted_hosts属性，允许数据和任意主机共享，默认不允许。.setuid_username（UID） gmetad设置UID的用户名。默认为nobody。.setuid（on|off） 当该值设置为off时，将不能设置UID,使用当前用户。.xml_port（数字） gmetad侦听端口(监听其他gmetad)，默认为8651。.umask（数字） 指定已创建RRD文件及其目录(/var/lib/ganglia/rrds)的umask。默认为022。.interactive_port（数字） gmetad交互式应答端口，默认为8652。 #暂时没有验证.server_threads（数字） 允许同时连接到侦听端口的连接数，默认为4。.case_sensitive_hostnames（|） 在gmetad之前版本中，RRD文件区分主机名大小写，但是现在已经有所改变。如果希望继续使用Ganglia .2之前版本创建的RRD文件，将该值设置为1。从Ganglia .2开始，该值默认设置为0。.RRA（文本格式） 该选项代表自定义的罗宾环存档（Round Robin Archive）值。默认为（15秒步进）： "RRA:AVERAGE:0.5:1:5856" "RRA:AVERAGE:0.5:4:20160" "RRA:AVERAGE:0.5:40:52704".rrd_rootdir（路径） 指定RRD文件在本地文件系统存储的基本目录（默认:/var/lib/ganglia/rrds）。

#!/usr/bin/env python# -*- coding: utf-8 -*-import randomimport timedef random_number_1(name): #ganglia调用此函数时会带上name实参,必须有一个行参。 return int(random.uniform(5, 50)) + 10def random_number_2(name): return int(random.uniform(50, 100)) + 10def metric_init(params): #必须带有一个参数 d1 = { #特定格式 'name': 'random_number_1', #度量名称,接下来的配置文件中调用的就是这个度量名称 'call_back': random_number_1, #在收集度量数据时被调用的函数 'time_max': 90, 'value_type': 'uint', #值类型，string|uint|float|double 'units': 'C', #度量单位，%|GB|MB|KB 'slope': 'both', 'format': '%u', #必须符合value_type，否则你的度量就会未知（参考: http://docs.python.org/library/stdtypes.html#string-formatting） 'description': 'Random a number', #度量的描述，在前台网页中，在划过主机度量图形时被呈现。 'groups': 'example random' #相同分组的度量在前台网页中会被关联在一起。 } d2 = { 'name': 'random_number_2', 'call_back': random_number_2, 'time_max': 90, 'value_type': 'uint', 'units': 'C', 'slope': 'both', 'format': '%u', 'description': 'Random a number', 'groups': 'example random' } descriptors = [d1,d2] #多个值使用列表返回 return descriptorsdef metric_cleanup(): pass

#!/usr/bin/env python# -*- coding: utf-8 -*-import osimport redef listen_port(name): count=0 with open('/proc/net/tcp','r') as f: for i in f: if not i.strip().startswith('s') and int(name) == int(re.split('[: ]+',i)[3],16): count+=1return 1 if count > 0 else 0def metric_init(params): descriptors = [] for i in params.values(): descriptors.append({ 'name': i, 'call_back': listen_port, 'time_max': 90, 'value_type': 'uint', 'units': 'units', 'slope': 'both', 'format': '%d', 'description': 'TCP_'+str(i), 'groups': 'Listen Port' }) return descriptorsdef metric_cleanup(): pass

modules { module { name = "listen_port" language = "python" path = "/usr/lib64/ganglia/python_modules/listen_port.py" param mysql_port { value = } param nginx_port { value = } param tomcat_port { value = }}}collection_group { collect_every = time_threshold = metric { name = #metric名称需要与value值相同。 title = "TCP_80" value_threshold = } metric { name = title = "TCP_3306" value_threshold = } metric { name = title = "TCP_8080" value_threshold = }}

[root@nagios python_modules]# vim captcha.py#!/usr/bin/env python# -*- coding: utf-8 -*-import osimport reimport copyimport urllib2captcha_address={}def get_captcha(name): url=captcha_address[name] response=urllib2.urlopen(url) return 1 if len(response.read()) > 100 else 0 #根据验证码的长度来返回指标def metric_init(params): global captcha_address captcha_address=copy.copy(params) descriptors = [] for i in params.keys(): descriptors.append({ 'name': i, 'call_back': get_captcha, 'time_max': 90, 'value_type': 'uint', 'units': 'units', 'slope': 'both', 'format': '%d', 'description': 'jinyunhui'+i, 'groups': 'captcha' }) return descriptorsdef metric_cleanup(): pass

[root@nagios python_modules]# vim /usr/local/ganglia/etc/conf.d/verification_code.confmodules { module { name = "captcha" language = "python" path = "/usr/local/ganglia/lib64/ganglia/python_module/captcha.py" param frond { value = 'http://jyh.xxx-cloud.com/api/admin/captcha.jpg' } param back { value = 'http://ht.xxx-cloud.com/api/admin/captcha.jpg' }}}collection_group { collect_every = time_threshold = metric { name = frond title = "frond captcha" value_threshold = } metric { name = back title = "back captcha" value_threshold = }}

[root@iZbp1bmji0bwyxg4zamhzhZ ~]# vim /etc/ganglia/conf.d/mysql.pyconfmodules { module { name = "mysql" language = "python" param user { value = root #数据库用户 } param passwd { value = #数据库密码}param host { value = localhost #连接的主机，默认}}} #具体有哪些参数可以到py文件中查看。collection_group { collect_every = time_threshold = metric { name = "mysql_threads_connected" #具体有那些metric,可以手动执行py文件来查看。 title = "ClientConnNum" #value_threshold = 1.0 }}

wget http://sourceforge.net/projects/mysql-python/files/mysql-python/1.2.3/MySQL-python-1.2.3.tar.gztar zxvf MySQL-python-1.2..tar.gzcd MySQL-python-1.2./python setup.py buildyum -y install python-develpython setup.py buildpython setup.py install

#vim /usr/lib64/ganglia/python_modules/Monitor_NetworkIO.py#!/usr/bin/env python# -*- coding: utf-8 -*-from aliyunsdkcore import clientfrom aliyunsdkcore.profile import region_providerfrom aliyunsdkcms.request.v20180308 import QueryMetricLastRequestdef NetIn(name): clt = client.AcsClient('AccessKeyID','AccessKeySecret','cn-hangzhou') request = QueryMetricLastRequest.QueryMetricLastRequest() request.set_accept_format('json') request.add_query_param('Project', 'acs_ecs_dashboard') request.add_query_param('Period', '') request.add_query_param('Dimensions', '{"instanceId":"xxx"}') request.add_query_param('Metric', 'VPC_PublicIP_InternetInRate') response = clt.do_action(request) return eval(eval(response.decode())['Datapoints'])[0]["Average"]def NetOut(name): clt = client.AcsClient('AccessKeyID','AccessKeySecret','cn-hangzhou') request = QueryMetricLastRequest.QueryMetricLastRequest() request.set_accept_format('json') request.add_query_param('Project', 'acs_ecs_dashboard') request.add_query_param('Period', '') request.add_query_param('Dimensions', '{"instanceId":"xxx"}') request.add_query_param('Metric', 'VPC_PublicIP_InternetOutRate') response = clt.do_action(request) return eval(eval(response.decode())['Datapoints'])[0]["Average"]def metric_init(params): d1={ 'name': 'NetIn', 'call_back': NetIn, 'time_max': 90, 'value_type': 'uint', 'units': 'units', 'slope': 'both', 'format': '%d', 'description': 'NetworkIO_NetIn', 'groups': 'NetworkIO' } d2={ 'name': 'NetOut', 'call_back': NetOut, 'time_max': 90, 'value_type': 'uint', 'units': 'units', 'slope': 'both', 'format': '%d', 'description': 'NetworkIO_NetIn', 'groups': 'NetworkIO' } descriptors=[d1,d2] return descriptorsdef metric_cleanup(): pass

modules { module { name = "Monitor_NetworkIO" language = "python" path = "/usr/lib64/ganglia/python_modules/Monitor_NetworkIO.py"}}collection_group { collect_every = time_threshold = metric { name = "NetIn" title = "NetworkIO_In" value_threshold = } metric { name = "NetOut" title = "NetworkIO_Out" value_threshold = }}

[root@nagios ~]# cat /usr/local/nagios/etc/objects/services.cfgdefine servicegroup { #定义相应的服务组 servicegroup_name ganglia-metrics alias Ganglia Metrics}define host { #定义监控的主机 use linux-server host_name node4 address 192.168.18.181}define host { use linux-server host_name node3 address 192.168.18.227}define host { use linux-server host_name aliyun_test address 47.98.117.8}define hostgroup { #定义监控组,包含监控的主机 hostgroup_name rimi-servers members node4,node3,aliyun_test}define command { #定义获取ganglia监控数据的命令 command_name check_ganglia command_line $USER1$/check_ganglia.py -h $HOSTADDRESS$ -m $ARG1$ -w $ARG2$ -c $ARG3$} #-h 为你定义的主机IP,-m 为获取主机的metric值，-w 为警告值，-c 为临界值。define service { use generic-service name ganglia-service hostgroup_name rimi-servers service_groups ganglia-metrics notifications_enabled }define service { use ganglia-service service_description load_one check_command check_ganglia!load_one!!}define service { use ganglia-service service_description cpu_idle check_command check_ganglia!cpu_idle!!}define service { use ganglia-service service_description part_max_used check_command check_ganglia!part_max_used!!}define service { #定义监控Nginx端口 use generic-service host aliyun_test service_description Nginx_Port check_command check_ganglia!!0.5!}define service { use generic-service host aliyun_test service_description MySQL_Port check_command check_ganglia!!0.5!}define service { #定义监控网卡流量 use generic-service host aliyun_test service_description NetIn check_command check_ganglia!NetIn!!}define service { use generic-service host aliyun_test service_description NetOut check_command check_ganglia!NetOut!!}define service { #定义监控MySQL指标 use generic-service host aliyun_test service_description mysql_threads_connected check_command check_ganglia!mysql_threads_connected!!}

#vim /usr/local/ganglia/etc/gmetad.conftrusted_hosts 192.168.18.188

#vim /usr/local/nagios/libexec/check_ganglia.pyganglia_host = '192.168.18.188'ganglia_port =

[root@nagios python_modules]# tail -20 /usr/local/nagios/libexec/check_ganglia.py if critical > warning: if value >= critical: print "CHECKGANGLIA CRITICAL: %s is %.2f" % (metric, value) sys.exit(2) elif value >= warning: print "CHECKGANGLIA WARNING: %s is %.2f" % (metric, value) sys.exit(1) else: print "CHECKGANGLIA OK: %s is %.2f" % (metric, value) sys.exit(0) else: if critical >= value: print "CHECKGANGLIA CRITICAL: %s is %.2f" % (metric, value) sys.exit(2) elif warning >= value: print "CHECKGANGLIA WARNING: %s is %.2f" % (metric, value) sys.exit(1) else: print "CHECKGANGLIA OK: %s is %.2f" % (metric, value) sys.exit(0)

[root@nagios ~]# vim /usr/local/ganglia/etc/gmond.confudp_send_channel { host = 192.168.18.213 port = ttl = } #发送数据到本机udp_recv_channel {
port = retry_bind = true bind = 192.168.18.213}#接收本机8649端口的数据

[root@gitlab ~]# vim /etc/ganglia/gmond.confglobals { daemonize = yes setuid = yes user = ganglia debug_level = max_udp_msg_len = mute = no deaf = yes #修改为不接收数据 allow_extra_data = yes host_dmax = /*secs. Expires (removes from web interface) hosts in 1 day */ host_tmax = /*secs */ cleanup_threshold = /*secs */ gexec = no send_metadata_interval = /*secs */ #定期的发送元数据。}udp_send_channel { host = 192.168.18.213 port = ttl = }

[root@ganglia_test ganglia]# vim /usr/local/ganglia/etc/gmetad.confgridname "AliyunGanglia" #这里要设置成其他的名称，多个gmetad之间名称不能相同。authority http://47.98.197.7/ganglia/ #修改选择Grid时，跳转的url。默认是主机名trusted_hosts 127.0.0.1 61.130.182.122 #让61.130.182.122拥有对8651端口的xml的访问权限

[root@nagios ganglia-3.7.]# vim /usr/local/ganglia/etc/gmetad.confdata_source "aliyunClient" 47.98.197.7: #添加监听子节点