# node_mgr

**Repository Path**: spkx/node_mgr

## Basic Information

- **Project Name**: node_mgr
- **Description**: 在每个机器上运行一个节点管理服务，以执行来自外部的命令，包括创建/启动/停止 存储或者计算节点，备份数据等功能
- **Primary Language**: C++
- **License**: Apache-2.0
- **Default Branch**: main
- **Homepage**: http://www.zettadb.com
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 1
- **Created**: 2025-02-11
- **Last Updated**: 2025-02-11

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

## node_mgr

Each machine node runs such an server instance, which operates kunlun storage nodes(SN) and/or computing nodes(CN) that run on the computer server, in response to requests initiated by computing nodes and/or cluster_mgr module to start/stop/create/drop/backup/restore a CN/SN. 


## the http post sample int the node_mgr/test/

http_post_para.cc
http_post_file.cc
./http_post_para http://127.0.0.1:9998 "{\"key\":\"value\"}"
./http_post_file http://127.0.0.1:9998 "{\"key\":\"value\"}" ./test.json

it can use curl, but curl can not send paramemter and file together.
curl -d "{\"ver\":\"0.1\",\"job_type\":\"get_node\",\"node_type\":\"meta_node\"}" http://127.0.0.1:9998/


## http post paramemter to cluster_mgr

1. get meta_node/storage_node/computer_node from cluster_mgr, with local ip :
{"ver":"0.1","job_type":"get_node","node_type":"meta_node","node_ip0":"127.0.0.1","node_ip1":"192.168.207.145"}

2. it can get all node without any node_type or node_ip :
{"ver":"0.1","job_type":"get_node"}


## http post paramemter to node_mgr

1. get node of meta_node/storage_node/computer_node from node_mgr.
{"ver":"0.1","job_type":"get_node","node_type":"meta_node"}

curl -d "{\"ver\":\"0.1\",\"job_type\":\"get_node\",\"node_type\":\"meta_node\"}" http://127.0.0.1:9998/
./http_post_para http://127.0.0.1:9998 "{\"ver\":\"0.1\",\"job_type\":\"get_node\",\"node_type\":\"meta_node\"}"

2. get info from node_mgr, it will wait a moment to get machine info
{"ver":"0.1","job_type":"get_info"}

curl -d "{\"ver\":\"0.1\",\"job_type\":\"get_info\"}" http://127.0.0.1:9998/
./http_post_para http://127.0.0.1:9998 "{\"ver\":\"0.1\",\"job_type\":\"get_info\"}"

3. update node info is notified by cluster_mgr, and node_mgr will get all node from cluster_mgr.
{"ver":"0.1","job_type":"update_node"}

4. delete a table or binlong
{"ver":"0.1","job_id":"id123456","job_type":"delete","file_type":"table",
"s_ip":"127.0.0.1","s_port":6007,"s_user":"pgx","s_pwd":"pgx_pwd","s_dbname":"postgres_$$_public","s_table_name":"t13"}

{"ver":"0.1","job_id":"id123456","job_type":"delete","file_type":"binlog",
"s_ip":"127.0.0.1","s_port":6007, "s_user":"pgx","s_pwd":"pgx_pwd"}

5. send a file to another mysql instance
{"ver":"0.1","job_id":"id123456","job_type":"send","file_type":"table",
"s_ip":"127.0.0.1","s_port":6004,"s_user":"pgx","s_pwd":"pgx_pwd","s_dbname":"postgres_$$_public","s_table_name":"t12",
"d_ip":"127.0.0.1","d_port":6007,"d_user":"pgx","d_pwd":"pgx_pwd","d_dbname":"postgres_$$_public","d_table_name":"t17"}

6. send a cmd for node to run in the path of mysql/pgsql/cluster_mgr
{"ver":"0.1","job_type":"mysql_cmd","cmd":"./startmysql.sh 6001"}
{"ver":"0.1","job_type":"pgsql_cmd","cmd":"python2 start_pg.py port=5401"}
{"ver":"0.1","job_type":"cluster_cmd","cmd":"bash -x bin/cluster_mgr_safe --debug --pidfile=run.pid cluster_mgr.cnf"}

curl -d "{\"ver\":\"0.1\",\"job_type\":\"mysql_cmd\",\"cmd\":\"./stopmysql.sh 6007\"}" http://127.0.0.1:9998
curl -d "{\"ver\":\"0.1\",\"job_type\":\"mysql_cmd\",\"cmd\":\"./startmysql.sh 6007\"}" http://127.0.0.1:9998
./http_post_para http://127.0.0.1:9998 "{\"ver\":\"0.1\",\"job_type\":\"mysql_cmd\",\"cmd\":\"./stopmysql.sh 6007\"}"
./http_post_para http://127.0.0.1:9998 "{\"ver\":\"0.1\",\"job_type\":\"mysql_cmd\",\"cmd\":\"./startmysql.sh 6007\"}"

7. get job status
{"ver":"0.1","job_id":"id123456","job_type":"get_status"}

8. stop auto pullup with time of minutes, before stop a node, otherwise node_mgr will pull up the node soon.
{"ver":"0.1","job_id":"id123456","job_type":"auto_pullup","minutes":10,"port":6001}
"minutes":-1	:  stop auto pullup always
"minutes":0	:  start auto pullup
"minutes":m	:  stop auto pullup in m minutes
if the json without the item "port", it will done on all of the port.

9. start cold backup, it will backup mysql and save to hdfs
{"ver":"0.1","job_id":"id123456","job_type":"coldbackup","s_ip":"127.0.0.1","s_port":6004,"s_user":"pgx","s_pwd":"pgx_pwd"}

10. start cold restore, it will restore mysql from hdfs
{"ver":"0.1","job_id":"id123456","job_type":"coldrestore","s_ip":"127.0.0.1","s_port":6004,"s_user":"pgx","s_pwd":"pgx_pwd",
"file":"/coldbackup/clust1/shard1/coldback_20211122_234303.tgz"}


## http post file to node_mgr, and save in the path of mysql/pgsql/cluster_mgr

1. send a file with paramemter for node_mgr to save
./http_post_file http://127.0.0.1:9998 "{\"ver\":\"0.1\",\"file_type\":\"mysql\"}" ./mysql_shard1.json
./http_post_file http://127.0.0.1:9998 "{\"ver\":\"0.1\",\"file_type\":\"pgsql\"}" ./postgresql_comp.json
./http_post_file http://127.0.0.1:9998 "{\"ver\":\"0.1\",\"file_type\":\"cluster\"}" ./cluster_mgr.cnf