Splunk tips

1、Splunk HF 转发syslog到第三方主机

outputs.conf

[syslog]
defaultGroup=syslogGroup

[syslog:syslogGroup]
type = udp
server = 10.203.19.22:514

props.conf

[sourcetype::cosmo_syslog]
TRANSFORMS-nyc = splunk_to_syslog

[sourcetype::cosmo_cisco:asa]
TRANSFORMS-nyc = splunk_to_syslog

[sourcetype::cosmo_cisco_asa_anquan]
TRANSFORMS-nyc = splunk_to_syslog

transforms.conf
[splunk_to_syslog]
REGEX = .
DEST_KEY = _SYSLOG_ROUTING
FORMAT = syslogGroup

2、splunkforwarder修改8089端口(管理端口)

2.1、修改web.conf配置文件

查看/opt/splunkforwarder/etc/system/local/web.conf文件是否存在

[root@linux_mysql local]# ll /opt/splunkforwarder/etc/system/local/
web.conf
-rw-r--r--. 1 root root 52101 Jul  4 00:07 /opt/splunkforwarder/etc/system/local/web.conf
[root@linux_mysql local]# 

如果web.conf文件不存在将../default/下面的web.conf 复制到local目录下

cd /opt/splunkforwarder/etc/system/local/
cp ../default/web.conf .

2.2、修改web.conf配置文件

找到mgmtHostPort配置项并修改

#   Version 7.2.5
# DO NOT EDIT THIS FILE!
# Changes to default files will be lost on update and are difficult to
# manage and support.
#
# Please make any changes to system defaults by overriding them in
# apps or $SPLUNK_HOME/etc/system/local
# (See "Configuration file precedence" in the web documentation).
#
# To override a specific setting, copy the name of the stanza and
# setting to the file where you wish to override it.
#
# This file contains possible attributes and values you can use to configure Splunk's web interface.
#
[default]

[settings]

# enable/disable the appserver
startwebserver = 1

# port number tag is missing or 0 the server will NOT start an http listener
# this is the port used for both SSL and non-SSL (we only have 1 port now).
httpport = 8000

# this determines whether to start SplunkWeb in http or https.
enableSplunkWebSSL = false

# location of splunkd; don't include http[s]:// in this anymore.
#mgmtHostPort = 127.0.0.1:8089--此项配置
mgmtHostPort = 127.0.0.1:8099

# list of ports to start python application servers on (although usually
# one port is enough)  Set to 0 to instead run the application server
# directly as the web front end on 'httpport', separate from splunkd.
appServerPorts = 8065

# default timeout, in seconds, when communicating with splunkd
splunkdConnectionTimeout = 30

# enable/disable custom netloc when using http client
enableSplunkWebClientNetloc = False

# SSL certificate files.
privKeyPath = $SPLUNK_HOME/etc/auth/splunkweb/privkey.pem
serverCert = $SPLUNK_HOME/etc/auth/splunkweb/cert.pem

2.3、重启splunkforwarder服务

/opt/splunkforwarder/bin/splunk start

2.4、查看端口是否修改成功

ps -ef|grep splunk
[root@linux_mysql local]# ps -ef|grep splunk
root       7698      1  0 00:10 ?        00:00:01 splunkd --under-systemd --systemd-delegate=yes -p 8099 _internal_launch_under_systemd
root       7723   7698  0 00:10 ?        00:00:00 [splunkd pid=7698] splunkd --under-systemd --systemd-delegate=yes -p 8099 _internal_launch_under_systemd [process-runner]
root       7949   5126  0 00:28 pts/0    00:00:00 grep --color=auto splunk
[root@linux_mysql local]# 

3、使用splunk sourcetype过滤prometheus指标项

问题：将普罗米修斯数据接入到splunk系统中发现数据量非常庞大(有1560个指标项),真正使用到的指标项不到100个，需要过滤掉不需要的指标

3.1、需要设置的服务器角色

在HF和indexer上设置

3.2、配置文件

props.conf

[prometheus_filter]
TIME_FORMAT = %s%3N
TIME_PREFIX = }\s[\d\-\.]+\s
TRANSFORMS-set = setnull,conform
NO_BINARY_CHECK = true
SHOULD_LINEMERGE = false
pulldown_type = 1
category = Metrics
EXTRACT-metric_name,num1,num2 = ^(?P<metric_name>[^\{]+)[^\}\n]*\}\s+(?P<num1>[^ ]+)\s+(?P<num2>.+)
EXTRACT-metric_name,num1,num2：提取metric_name,num1,num2字段 TIME_FORMAT,TIME_PREFIX:时间格式 TRANSFORMS-set:过滤条件

transforms.conf

[conform]
REGEX = container_cpu_load_average_10s|container_cpu_usage_seconds_total|container_fs_limit_bytes|container_fs_usage_bytes|container_memory_cache|container_memory_usage_bytes|container_memory_working_set_bytes|container_network_receive_bytes_total|container_network_transmit_bytes_total|container_spec_memory_limit_bytes|jvm_gc_collection_seconds_count|jvm_gc_collection_seconds_sum|jvm_memory_bytes_used|jvm_memory_pool_bytes_max|jvm_memory_pool_bytes_used|jvm_threads_current|jvm_threads_deadlocked|jvm_threads_peak|kube_pod_container_status_ready|kube_pod_container_status_running|kube_pod_container_status_terminated|kube_pod_container_status_waiting|machine_cpu_cores|machine_memory_bytes|mysql_global_status_aborted_clients|mysql_global_status_commands_total|mysql_global_status_innodb_data_fsyncs|mysql_global_status_innodb_data_reads|mysql_global_status_innodb_data_writes|mysql_global_status_select_full_join|mysql_global_status_select_scan|mysql_global_status_slow_queries|mysql_global_status_sort_scan|mysql_global_status_threads_connected|mysql_global_status_threads_created|nginx_connections_accepted|nginx_connections_active|nginx_connections_handled|nginx_http_requests_total|node_network_info|rabbitmq_channelsTotal|rabbitmq_connectionsTotal|rabbitmq_consumersTotal|rabbitmq_exchangesTotal|rabbitmq_fd_total|rabbitmq_fd_used|rabbitmq_queue_messages_ready_total|rabbitmq_queue_messages_total|rabbitmq_queue_messages_unacknowledged_total|rabbitmq_queuesTotal|rabbitmq_sockets_total|rabbitmq_sockets_used|rabbitmq_up|redis_blocked_clients|redis_commands_processed_total|redis_connected_clients|redis_evicted_keys_total|redis_expired_keys_total|redis_keyspace_hits_total|redis_keyspace_misses_total|redis_memory_used_bytes|zk_max_latency|zk_num_alive_connections|zk_open_file_descriptor_count|zk_outstanding_requests|zk_up|zk_watch_count|zk_znode_count
DEST_KEY = queue
FORMAT = indexQueue

[setnull]
REGEX = .
DEST_KEY = queue
FORMAT = nullQueue

4、nix add-on CPU.sh 问题

问题：nix ad-on下发到Linux 系统上没有采集到cpu性能数据
需要在linux系统上安装以下命令

yum install sysstat
yum install sar
yum insatll mpstat

5、forwarder被克隆，导致主机名一致

5.1、解决方式一:

删除  $SPLUNK$/etc/instance.cfg  (里面记录了guid，也就是客户端名称)
删除server.conf里面的实例名称记录，inputs.conf里面的主机名称记录
重启forwarder

5.2、解决方式二:

forwarder重新安装
例如： 安装目录为/tpdata

/tpdata/splunkforwarder/bin/splunk stop
rm -rf  /tpdata/splunkforwarder
tar zxvf /tpdata/splunkforwarder-6*.tgz -C /tpdata
/tpdata/splunkforwarder/bin/splunk start --answer-yes --no-prompt --accept-license
/tpdata/splunkforwarder/bin/splunk set deploy-poll 10.21.8.228:8089 -auth admin:changeme
/tpdata/splunkforwarder/bin/splunk restart

6、修改每个CPU的最大并发搜索次数

limits.conf
[search] max_searches_per_cpu
# 每个CPU的最大并发搜索次数

7、关于Splunk 重启日志也会存放在index里面

index="_internal" source="/opt/splunk/var/log/splunk/splunkd.log" start

8、查看告警邮箱是否发送

INFO：为正常发送，ERROR：为发送失败  >>要检查ERROR报错的问题
index ="_internal" source="/opt/splunk/var/log/splunk/python.log" "mail"

9、关于字段重命名注意点

字段重命名：不能设置多个字段名称重命名为同一个字段名称

重写字段值：如果有设置多个字段名称重命名为同一个字段名称，使用勾选此选项字段重命名。

10、仪表板设置隐藏图表（并传递传递参数）

<row>
    <panel> depends="$range$"
        <table>
            <title>僵尸进程详情</title>
        </table>
    </panel>
</row>

11、Splunk web刷新配置文件url

http://192.168.10.9:8000/zh-CN/debug/refresh

12、修改index存放路径

vim /opt/splunk/splunk/etc/splunk-launch.conf --编辑splunk-launch文件

cat /opt/splunk/splunk/etc/splunk-launch.conf

#   Version 7.2.5

# Modify the following line to suit the location of your Splunk install.
# If unset, Splunk will use the parent of the directory containing the splunk
# CLI executable.
#
# SPLUNK_HOME=/opt/splunk-home
# By default, Splunk stores its indexes under SPLUNK_HOME in the
# var/lib/splunk subdirectory.  This can be overridden
# here:
#SPLUNK_DB=/opt/splunk-home/var/lib/splunk
#添加以下内容
SPLUNK_DB=/opt/splunk/splunk/data

# Splunkd daemon name
SPLUNK_SERVER_NAME=Splunkd

# Splunkweb daemon name
SPLUNK_WEB_NAME=splunkweb

# If SPLUNK_OS_USER is set, then Splunk service will only start
# if the 'splunk [re]start [splunkd]' command is invoked by a user who
# is, or can effectively become via setuid(2), $SPLUNK_OS_USER.
# (This setting can be specified as username or as UID.)
# SPLUNK_OS_USER

mkdir data --创建data目录
splunk stop --停止splunk服务
mv /opt/splunk/splunk/var/lib/splunk /opt/splunk/splunk/data/
将/opt/splunk/splunk/var/lib/splunk目录下的文件移动到/opt/splunk/splunk/data/目录下
splunk start --启动服务

13、Splunk License 安装位置

$SPLUNK_HOME/etc/licenses/enterprise/Splunk.License.lic

14、outputlookup小技巧

index=wineventlog 
| lookup AXA_Server_List Segment as 源网络地址 OUTPUT 分公司区域 机构 省份 Region 
| search Region=* 
| stats count by 分公司区域 省份 Region 源网络地址 username 
| search username!="-" username!="ANONYMOUS LOGON" username!=admin* 
| rename 源网络地址 as IP username as nt_host 
| rex field=nt_host "(?P<nt_host>[^$]*)" 
| eval nt_host=upper(nt_host) 
| lookup AXA_ServerIP_List.csv IP nt_host OUTPUT IP as ip 
| search NOT ip=* 
| fields - ip count 
| outputlookup AXA_ServerIP_List.csv append=true