1. 程式人生 > >Nagios監控mongodb分片集群服務實戰

Nagios監控mongodb分片集群服務實戰

check 演示 database 路由 easy interval oca mis har



1,監控插件下載
Mongodb插件下載地址為:git clone git://github.com/mzupan/nagios-plugin-mongodb.git,剛開始本人這裏沒有安裝gitpub環境,找網友草根幫忙下載的。之後上傳到了csdn資源頁面,新的下載地址為:http://download.csdn.net/detail/mchdba/8019077


2,加入新的mongodb監控命令

由於mongodb服務是和mysql從庫公用一臺物理機。之前已經做了基礎nagios以及mysql服務監控,所以這裏僅僅須要在原來的基礎上加入mongodb命令和服務就可以。Nagios監控mysql請參考:http://blog.itpub.net/26230597/viewspace-760141/以及http://blog.itpub.net/26230597/viewspace-1217246/。所以這裏須要加入的mongodb監控命令例如以下所看到的:


[[email protected] objects]# cd /usr/local/nagios/etc/objects
[[email protected] objects]# vim commands.cfg
define command {
    command_name check_mongodb
    command_line $USER1$/nagios-plugin-mongodb/check_mongodb.py -H $HOSTADDRESS$ -A $ARG1$ -P $ARG2$ -W $ARG3$ -C $ARG4$
}

define command {
    command_name check_mongodb_database
    command_line $USER1$/nagios-plugin-mongodb/check_mongodb.py -H $HOSTADDRESS$ -A $ARG1$ -P $ARG2$ -W $ARG3$ -C $ARG4$ -d $ARG5$
}

define command {
    command_name check_mongodb_collection
    command_line $USER1$/nagios-plugin-mongodb/check_mongodb.py -H $HOSTADDRESS$ -A $ARG1$ -P $ARG2$ -W $ARG3$ -C $ARG4$ -d $ARG5$ -c $ARG6$
}

define command {
    command_name check_mongodb_replicaset
    command_line $USER1$/nagios-plugin-mongodb/check_mongodb.py -H $HOSTADDRESS$ -A $ARG1$ -P $ARG2$ -W $ARG3$ -C $ARG4$ -r $ARG5$
}

define command {
    command_name check_mongodb_query
    command_line $USER1$/nagios-plugin-mongodb/check_mongodb.py -H $HOSTADDRESS$ -A $ARG1$ -P $ARG2$ -W $ARG3$ -C $ARG4$ -q $ARG5$
}


3,加入mongodb監控服務mongodb的服務也須要單獨又一次加入。例如以下所看到的:
#檢測mongodb服務的連接時間,超過2秒就普通報警,5秒就嚴重報警
define service{
        host_name dbm1slave1
        service_description Mongo Connect Check
        check_command check_mongodb!connect!30000!2!5
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }

#檢查mongodb的連接數,超過150普通報警,200嚴重報警
define service{
        host_name dbm1slave1
        service_description Mongo Free Connections
        check_command check_mongodb!connections!27017!70!80
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
        
#檢查mongodb復制完畢的百分比率,確保primary和standby的time是一致的。
define service{
        host_name dbm1slave1
        service_description Mongo Replication Lag
        check_command check_mongodb!replication_lag!27017!15!30
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
#檢查mongodb內存使用率。閥值與mongodb所在機器的總內存數相關
define service{
        host_name dbm1slave1
        service_description Mongo Memory Usage
        check_command check_mongodb!memory!27017!20!28
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
#檢查mongodb Mapped的內存使用率。閥值與mongodb所在機器的總內存數相關
define service{
        host_name dbm1slave1
        service_description Mongo Mapped Memory Usage
        check_command check_mongodb!memory_mapped!27017!20!28
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
#檢查Lock Time的百分率。假設lock time占領mongo運行時間的5%就普通報警。假設超過10%就嚴重報警
define service{
        host_name dbm1slave1
        service_description Mongo Lock Percentage
        check_command check_mongodb!lock!27017!5!10
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }

# Check Average Flush Time,檢查mongo服務器的平均flush時間,
define service{
        host_name dbm1slave1
        service_description Mongo Flush Average
        check_command check_mongodb!flushing!27017!100!200
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }

# Check Last Flush Time,檢查最新的flush時間,假設超過200ms就普通報警。超過400ms就嚴重報警
define service{
        host_name dbm1slave1
        service_description Mongo Last Flush Time
        check_command check_mongodb!last_flush_time!27017!200!400
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
# Check status of mongodb replicaset,檢查mongo復制的狀態
define service{
        host_name dbm1slave1
        service_description MongoDB state
        check_command check_mongodb!replset_state!27017!0!0
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }

# Check status of index miss ratio,檢查索引命中率。
define service{
        host_name dbm1slave1
        service_description MongoDB Index Miss Ratio
        check_command check_mongodb!index_miss_ratio!27017!.005!.01
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
# Check number of databases and number of collections
define service{
        host_name dbm1slave1
        service_description MongoDB Number of databases
        check_command check_mongodb!databases!27017!300!500
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
define service{
        host_name dbm1slave1
        service_description MongoDB Number of collections
        check_command check_mongodb!collections!27017!300!500
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }        
        
# Check size of a database,檢查庫的大小
define service{
        host_name dbm1slave1
        service_description MongoDB Database size your-database
        check_command check_mongodb_database!database_size!27017!300!500!your-database
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }                
        
# Check index size of a database,檢查庫索引的大小
define service{
        host_name dbm1slave1
        service_description MongoDB Database index size your-database
        check_command check_mongodb_database!database_indexes!27017!50!100!your-database
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }            
        
# Check index size of a collection,檢查集合collection的索引大小
define service{
        host_name dbm1slave1
        service_description MongoDB Database index size your-database
        check_command check_mongodb_collection!collection_indexes!27017!50!100!your-database!your-collection
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
# Check the primary server of replicaset。檢查復制的primary服務
define service{
        host_name dbm1slave1
        service_description MongoDB Replicaset Master Monitor: your-replicaset
        check_command check_mongodb_replicaset!replica_primary!27017!0!1!your-replicaset 
        #演示樣例:check_command check_mongodb_replicaset!replica_primary!27017!0!1!shard2
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }

        
# Check the number of queries per second,檢查每一秒的查詢數量
define service{
        host_name dbm1slave1
        service_description MongoDB Updates per Second
        check_command check_mongodb_query!queries_per_second!27017!200!150!update
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }
        
# Check Primary Connection,檢查復制中與primary庫的連接時間,超過2秒就普通報警,超過4秒就嚴重報警
define service{
        host_name dbm1slave1
        service_description Mongo Connect Check
        check_command check_mongodb!connect_primary!27017!2!4
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }

# Check Collection State。檢查collection狀態,檢查mongo服務組列表的每個主機,能夠檢查重要collection的高可用性(鎖、超時、服務配置的可用性)。假設發現一個查詢失敗就會報警。
define service{
        host_name dbm1slave1
        service_description Mongo Collection State
        check_command check_mongodb!collection_state!27017!your-database!your-collection
        max_check_attempts 5
        normal_check_interval 3
        retry_check_interval 2
        check_period 24x7
        notification_interval 10
        notification_period 24x7
        notification_options w,u,c,r
        contact_groups ops
        }




4,查看部分監控項效果

配置完nagios端服務。重新啟動下service nagios restart; 等上幾分鐘,nagios監控界面就會出現完整的mongo服務信息,例如以下所看到的:

技術分享


5
,從ps中確定mongodb的架構

[[email protected] ~]# ps -eaf|grep mongo

mongodb 2457 1 0 2013 ?

2-03:39:08 ./mongod --configsvr --dbpath /home/data/mongodb/config --port 20000 --logpath /home/data/mongodb/config.log --logappend --fork

mongodb 2804 1 0 2013 ? 1-10:02:33 mongos --configdb 192.168.12.62:20000,192.168.12.63:20000,192.168.12.72:20000 --port 30000 --chunkSize 64 --logpath /home/data/mongodb/mongos.log --logappend --fork

mongodb 3072 1 0 2013 ?

1-10:17:20 mongod --shardsvr --replSet shard1 --port 27017 --dbpath /home/data/mongodb/shard11 --oplogSize 2048 --logpath /home/data/mongodb/shard11.log --logappend --fork

root 11179 9391 0 11:14 pts/1 00:00:00 grep mongo

mongodb 30414 1 0 Feb14 ? 1-06:20:50 mongod --shardsvr --replSet shard2 --port 27018 --dbpath /home/data/mongodb/shard21 --oplogSize 2048 --logpath /home/data/mongodb/shard21.log --logappend --fork

[[email protected] ~]#

看到有4個mongo進程,

a) 啟動參數有“--configdb”的就是集群入口進程;

b) Shard Server,啟動參數帶“--shardsvr --replSet”的是集群分片的一個片組啟動進程,用戶存儲實際的數據塊,也就是27017port和27018port的mongodb服務實例。至於怎樣推斷27017port中哪個是primary哪個是secondary須要去登錄27107port運行rs.status();去查看一下。

c) Config Server:啟動參數帶“--configsvr”的進程,存儲了整個Cluster Metadata,當中包含chunk信息,也就是20000port的mongodb服務實例。

d) Route Server:啟動參數帶“mongos --configdb”的進程,前端路由,client由此接入。且讓整個集群看上去像單一數據庫,前端應用能夠透明使用。也就是30000port的mongodb實例。



6,調試中出現過的錯誤

錯誤1:

[[email protected] nagios ~]# tail -f /usr/local/nagios/var/nagios.log

[1412819956] Warning: Return code of 13 for check of service ‘Mongo Memory Usage‘ on host ‘dbm1slave1‘ was out of bounds.

[1412819956] SERVICE ALERT: dbm1slave1;Mongo Memory Usage;CRITICAL;SOFT;1;(Return code of 13 is out of bounds)

[1412819975] Warning: Return code of 13 for check of service ‘Mongodb Connect Check‘ on host ‘dbm1slave1‘ was out of bounds.

[1412819975] SERVICE ALERT: dbm1slave1;Mongodb Connect Check;CRITICAL;SOFT;1;(Return code of 13 is out of bounds)

[1412820058] Warning: Return code of 13 for check of service ‘Mongo Free Connections‘ on host ‘dbm1slave1‘ was out of bounds.

須要賦值nagios用戶全部權限以及r運行權限

chmod 770 /usr/lib/nagios/plugins/check_mongodb.py

chown -R nagios.nagios /usr/lib/nagios/plugins/check_mongodb.py

錯誤2:

監控界面Status Information一欄出現 No module named pymongo報錯提示信息:

出現這個提示是由於須要安裝pymongo模塊,運行easy_install pymongo命令安裝就可以。例如以下所看到的:

[[email protected] objects]# easy_install pymongo

Searching for pymongo

Reading http://pypi.python.org/simple/pymongo/

Best match: pymongo 2.7.2

......

zip_safe flag not set; analyzing archive contents...

Adding pymongo 2.7.2 to easy-install.pth file

Installed /usr/lib/python2.6/site-packages/pymongo-2.7.2-py2.6-linux-x86_64.egg

Processing dependencies for pymongo

Finished processing dependencies for pymongo

技術分享----------------------------------------------------------------------------------------------------------------

<版權全部,文章同意轉載,但必須以鏈接方式註明源地址,否則追究法律責任!>
原博客地址:http://blog.itpub.net/26230597/viewspace-1293589/
原作者:黃杉 (mchdba)

技術分享----------------------------------------------------------------------------------------------------------------


參考文章:https://github.com/mzupan/nagios-plugin-mongodb/blob/master/README.md


Nagios監控mongodb分片集群服務實戰