jpuyy的网志

Blog

  • 判断 zookeeper 健康状态

    echo stat | nc localhost 2181 | grep Mode
    You will get the response if the node is acting as a leader:

    Mode: leader
    If it is a follower, you will see:

    Mode: follower
    Alternatively, you can use the zkServer.sh located in /opt/zookeeper/bin:

    ./zkServer.sh status

    [root@misc010-prod ~]# echo mntr |nc  10.16.1.60 2181
zk_version	3.4.10-39d3a4f269333c922ed3db283be479f9deacaa0f, built on 03/23/2017 10:13 GMT
zk_avg_latency	0
zk_max_latency	2
zk_min_latency	0
zk_packets_received	36275
zk_packets_sent	36274
zk_num_alive_connections	2
zk_outstanding_requests	0
zk_server_state	leader
zk_znode_count	488
zk_watch_count	3
zk_ephemerals_count	4
zk_approximate_data_size	40646
zk_open_file_descriptor_count	33
zk_max_file_descriptor_count	65536
zk_followers	2
zk_synced_followers	2
zk_pending_syncs	0

[root@misc010-prod ~]# echo mntr |nc  10.16.1.21 2181
zk_version	3.4.10-39d3a4f269333c922ed3db283be479f9deacaa0f, built on 03/23/2017 10:13 GMT
zk_avg_latency	0
zk_max_latency	10
zk_min_latency	0
zk_packets_received	685
zk_packets_sent	686
zk_num_alive_connections	1
zk_outstanding_requests	0
zk_server_state	follower
zk_znode_count	488
zk_watch_count	0
zk_ephemerals_count	4
zk_approximate_data_size	40646
zk_open_file_descriptor_count	30
zk_max_file_descriptor_count	1048576
[root@misc010-prod ~]# echo mntr |nc  10.16.1.22 2181
zk_version	3.4.10-39d3a4f269333c922ed3db283be479f9deacaa0f, built on 03/23/2017 10:13 GMT
zk_avg_latency	0
zk_max_latency	13
zk_min_latency	0
zk_packets_received	73614
zk_packets_sent	73622
zk_num_alive_connections	3
zk_outstanding_requests	0
zk_server_state	follower
zk_znode_count	488
zk_watch_count	37
zk_ephemerals_count	4
zk_approximate_data_size	40646
zk_open_file_descriptor_count	32
zk_max_file_descriptor_count	1048576

    在 leader 输出中
    zk_followers 2
    zk_synced_followers 2
    zk_pending_syncs 0
    表示 zk 正常

  • kafka 查看 group id 和 offset

    查看所有 group id

     bin/kafka-consumer-groups.sh --bootstrap-server broker1:9092 --list

    描述具体的 group id

     bin/kafka-consumer-groups.sh --bootstrap-server broker1:9092 --describe --group test-consumer-group

  • Padavan 定时任务

    每天重启路由器，每周升级

    30 3 * * * reboot
# Upgrade every sunday
00 2 0 * * rm -f /tmp/up.sh; wget https://opt.cn2qq.com/opt-script/up.sh -P /tmp/; sh -x /tmp/up.sh

  • 判断一台服务器在不在用

    1. 看进程

    ps aux  | grep -Ei "java|rabbitmq|mongo|mysql|redis|nginx|python|php|zabbix|haproxy|keepalive|dnsmasq|bind|docker|httpd|oracle"

    2. 看连接

    netstat -ntp | grep -vE ":25|:22"

    3. 看所有进程及从属关系

    pstree -p

  • es 调整每个 node 最大 shard 数量

    新集群 5 台机器

    Validation Failed: 1: this action would add [2] total shards, but this cluster currently has [4999]/[5000] maximum shards open;

    默认每个node 1000个，则最多 5000 个 shard，满了导致不能创建新的 es index

    {
  "persistent" : {
    "cluster" : {
      "max_shards_per_node" : "100000"
    },
    "search" : {
      "max_buckets" : "50000"
    }
  },
  "transient" : {
    "cluster" : {
      "max_shards_per_node" : "100000"
    },
    "search" : {
      "max_buckets" : "50000"
    }
  }
}

  • 挂载远程 sftp 目录至 mac OSX 本机目录

    安装所需 osxfuse

    brew cask install osxfuse
brew install sshfs

    把远程目录 /data/v 挂载到 ~/v 目录下

    sshfs -p 22 -o reconnect -o volname=v -o IdentityFile=~/.ssh/id_rsa [email protected]:/path/to/wherever /Volumes/SomeNiceName

    参考:

    SSHFS mount remote drive in finder