check_mk raw 1.2.8p17 FAQ

Q：有没有已经实施的案例
A：
http://wiki.lustre.org/Check_MK/Graphite/Graphios_Setup_Guide

Q：check-mk-agent怎么安装？
A：
使用epel源，yum -y install check-mk-agent xinetd && /etc/init.d/xinetd start && chkconfig xinetd on，同时开放tcp:6556端口，目前epel源最高版本 1.2.6p16-3.el6

Q：客户端获取应用监控信息，如mysql等
A:
放置在/usr/share/check-mk-agent/plugins目录，如果需要异步的话，在plugins目录新建以秒为名称的目录，如/usr/share/check-mk-agent/plugins/60

Q：服务端过滤无用监控信息
A：
编辑/opt/omd/sites/monitor/etc/check_mk/main.mk，添加以下内容
ignored_checktypes = ["chrony", "kernel", "cpu.threads", "logwatch", "megaraid_pdisks", "ipmi", "ipmi_sensors", "mounts", "ps", "ps.perf", "postfix_mailq", "postfix_mailq_status", "logins", "omd_apache", "omd_status", "livestatus_status"]

Q：监控数据量大的时候，是如何做优化的
A：
单机用目录方式进行优化，另外多机的话，还支持分布式

Q：如何变更检测时间及重试次数
A：
Host & Service Parameters -> Monitoring Configuration -> Service Checks -> Normal check interval for service checks 设置1分钟
Host & Service Parameters -> Monitoring Configuration -> Service Checks -> Retry check interval for service checks 设置1分钟
Host & Service Parameters -> Monitoring Configuration -> Service Checks -> Maximum number of check attempts for service 设置1次

Host & Service Parameters -> Monitoring Configuration -> Service Checks -> Normal check interval for host checks 设置1分钟
Host & Service Parameters -> Monitoring Configuration -> Service Checks -> Retry check interval for host checks 设置1分钟
Host & Service Parameters -> Monitoring Configuration -> Service Checks -> Maximum number of check attempts for service 设置1次

Q：如何自动添加发现的新监控项
A：
Host & Service Parameters -> Monitoring Configuration -> Inventory and Check_MK settings -> Periodic service discovery 增加rule
Perform service discovery every 设置为1h
Severity of unmonitored services 设置为warning
Severity of vanished services 设置为ok
Automatically update service configuration中，mode设置为Add unmonitored services & remove varnished services
Group discovery and activation for up to 设置为10m

Q：支持批量主机导入吗
A：Host -> Bulk host import

Q：建立主机组并关联主机
A:
Host & Service Groups  -> new host group
Host Tags -> new tag group
Host & Service Parameters -> Grouping -> Assignment of hosts to host groups
新建rule ,  Assignment of hosts to host groups 选择group_bj，Conditions下host tag选择 location is group_bj

Q：磁盘io模式由summary变更为separate
A：
Host & Service Parameters -> Parameters for discovered services -> Storage, Filesystems and Files -> Discovery mode for Disk IO check
新建rule，选择Create a separate check for each physical disk

Q：网络设备接口名称由id变更为description
A：
Host & Service Parameters -> Parameters for discovered services -> Discovery - automatic service detection -> Network Interface and Switch Port Discovery，新建rule，选择Use description as service name for network interface checks，选择use description

Q：改变监控阀值
A：
#load
Host & Service Parameters -> Parameters for discovered services -> Operating System Resources -> CPU load (not utilization!)，添加rule，选择fixed level，warning & critical设置为10

#cpu
Host & Service Parameters -> Parameters for discovered services -> Operating System Resources -> CPU utilization on Linux/UNIX，添加rule，选择Alert on too high CPU utilization，warning & critical设置为95，选择Alert on too high disk wait (IO wait)，warning & critical设置为50

#memory
Host & Service Parameters -> Parameters for discovered services -> Operating System Resources -> Memory and Swap usage on Linux，添加rule，选择Levels for RAM，warning & critical设置为95

#storage
Host & Service Parameters -> Parameters for discovered services -> Storage, Filesystems and Files -> Filesystems (used space and growth)
添加rule，选择Levels for filesystem free space，dynamic level，
Filesystem larger than 1T -> Absolute free space -> warning & critical设置为20000MB
Filesystem larger than 300G -> Absolute free space -> warning & critical设置为10000MB
Filesystem larger than 100G -> Absolute free space -> warning & critical设置为5000MB
Filesystem larger than 10G -> Absolute free space -> warning & critical设置为1000MB
Filesystem larger than 1G -> Absolute free space -> warning & critical设置为100MB
Filesystem larger than 300M -> Absolute free space -> warning & critical设置为20MB
Filesystem larger than 100M -> Absolute free space -> warning & critical设置为10MB
Filesystem larger than 0B -> Absolute free space -> warning & critical设置为0MB

#traffic
Host & Service Parameters -> Parameters for discovered services -> Networking -> Network interfaces and switch ports，新建rule，

1g
Operating speed -> ignore speed，Operational state -> 1 - up，Assumed input speed -> 1g, Assumed output speed -> 1g， Measurement unit -> Bits，Used bandwidth (minimum or maximum traffic) -> in/out & upper & percent level & warning 95 & critical 95，Average values -> 10mins，Port Specification -> em1 em2 em3 em4 eth0 eth1 eth2 eth3

2g
Operating speed -> ignore speed，Operational state -> 1 - up，Assumed input speed -> 2000000000 bit, Assumed output speed -> 2000000000 bit， Measurement unit -> Bits，Used bandwidth (minimum or maximum traffic) -> in/out & upper & percent level & warning 95 & critical 95，Average values -> 10mins，Port Specification -> bond0 bond1

10g
Operating speed -> ignore speed，Operational state -> 1 - up，Assumed input speed -> 10g, Assumed output speed -> 10g， Measurement unit -> Bits，Used bandwidth (minimum or maximum traffic) -> in/out & upper & percent level & warning 95 & critical 95，Average values -> 10mins，Port Specification -> p1p1 p1p2 p2p1 p2p2

cmk -L |grep redis
redis.info                                   tcp    (no man page present)

cmk --checks=redis.info -I kvm-48-113

cmk --debug -nv kvm-48-113