Zabbix高级应用 --K8s集群监控 ����� ��������FiberHome 演讲主题： • CactiFans • Zabbix • Microservice • K8s • Golang https://blog.cactifans.com/ 无边界监控 应用架构发展趋势 数据采集 RESTful 接口 LLD – 低级别发现 低级发现提供了一种监控主机上变化实体，并 对采集器采集的原始数据通过管道进行处理，最终达到 数据展现和存储的要求。 例如 SNMP采集交换机内存 15MB string? 图形？触发器？ Kubernetes 架构 k8s node k8s pod k8s master 应用 Zabbix Agent HTTP Agent CPU/内存/磁盘/网络 服务/日志/... 状态/CPU/内存 Kubernetes 监控 Kubernetes

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238 2 Proxy v.s. Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 5 Configuration of a single Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 6 Switching feature might be very handy for automatic monitoring of new Cloud nodes. As soon as you have a new node in the Cloud Zabbix will automatically start collection of performance and availability data of the

process that may collect data on behalf of Zabbix server, taking some processing load off of the server node - a full Zabbix server configured as an element within a hierarchy of distributed monitoring; it are calculated in a similar way, but this also depends on a total number of child nodes linked to a node. Time synchronisation It is very important to have precise system date on server with Zabbix running Parameter Description Group name Enter a unique host group name. The name must be unique within a Zabbix node. Hosts Select hosts, members of the group. A host group may have zero, one or more hosts. 2 Inventory

• zabbix[java„] • zabbix[process,,,] (alerter, db watchdog, escalator, node watcher, proxy poller, timer processes aren’t supported, but two new processes are supported - data usage in environments with huge numbers of problem triggers and host groups. • In node-based environments, a duplicating Node column has been removed from trigger popups in the System status frontend widget process that may collect data on behalf of Zabbix server, taking some processing load off of the server node - a full Zabbix server configured as an element within a hierarchy of distributed monitoring; it

zabbix_server instances or nodes, where only one node can be active (working) at a time, while other nodes are on standby, ready to take over in case the current node is stopped or fails. See also: High availability now printed to the console. • It is possible to send longer input parameters, such as HA node name instead of node number. Runtime controls on BSD-based OS 15 Previously, Zabbix server and Zabbix proxy CockroachDB by HTTP • Envoy Proxy by HTTP • HashiCorp Consul Cluster by HTTP • HashiCorp Consul Node by HTTP See setup instructions for HTTP templates. You can get these templates: • In Configuration

status. ha_remove_node=target Remove the high availability (HA) node specified by its name or ID. Note that active/standby nodes cannot be removed. target - name or ID of the node (can be obtained high availability Starting Zabbix server as cluster node Two parameters are required in the server configuration to start a Zabbix server as cluster node: • HANodeName parameter must be specified for each each Zabbix server that will be an HA cluster node. This is a unique node identifier (e.g. zabbix-node-01) that the server will be referred to in agent and proxy configurations. If you do not specify

status. ha_remove_node=target Remove the high availability (HA) node specified by its name or ID. Note that active/standby nodes cannot be removed. target - name or ID of the node (can be obtained high availability Starting Zabbix server as cluster node Two parameters are required in the server configuration to start a Zabbix server as cluster node: • HANodeName parameter must be specified for each each Zabbix server that will be an HA cluster node. This is a unique node identifier (e.g. zabbix-node-01) that the server will be referred to in agent and proxy configurations. If you do not specify

status. ha_remove_node=target Remove the high availability (HA) node specified by its name or ID. Note that active/standby nodes cannot be removed. target - name or ID of the node (can be obtained availability Starting Zabbix server as cluster node 30 Two parameters are required in the server configuration to start a Zabbix server as cluster node: • HANodeName parameter must be specified for for each Zabbix server that will be an HA cluster node. This is a unique node identifier (e.g. zabbix-node-01) that the server will be referred to in agent and proxy configurations. If you do not specify

Output:\2 测试结果：10000 03 现场练习 现场练习 做一个Prometheus 的主从监控项 现场练习 Node-export 确认 1.登录测试机器 2.启动export 命令： /home/zabbix/node_exporter/node_exporter & 3. Curl http://127.0.0.1:9100/metrics 现场练习 监控项类型：http url：http://127.0.0.1:9100/metrics 数据类型：文本/text 现场练习 数据效果 现场练习 预处理类型：Prometheus pattern 预处理语法： node_filesystem_avail_bytes{device="/dev/vda1",fstype="ext4",mountpoint="/"} 04 一些备注与提问环节 主从监控项 Dependent

CockroachDB by HTTP • Envoy Proxy by HTTP • HashiCorp Consul Cluster by HTTP • HashiCorp Consul Node by HTTP 有关设置说明，请参阅HTTP 模板. 你可以使用如下方式获取到新的模板: • 在新安装的 配置 → 模板中; • 如果要从以前的版本升级，可以从 Zabbix Git repository 警报管理器统计 信息 lld - LLD 管理器统计信息 locks - 互斥锁列表（在 **BSD 系统 * 上为空） ha_status 记录高可用性 (HA) 集群状态。 ha_remove_node=target 删除由其列出的编号指定的高可用性 (HA) 节点。 请注意，无法删除活动/备用节点。 target - 列表中的节点编号 （可以通过运行 ha_status 获 得） ha_s • HANodeName 集群中每个 zabbix server 节点名称必须要有且唯一。 36 这是将在 agent 和 proxy 配置中引用 server 的名称 (例如：zabbix-node-01)。如果您没有指定 HANodeName，那么服务器将以 standalone 模式启动。 • NodeAddress 参数必须为每个节点指定。 NodeAddress 参数 (address:port)