Skywalking python agent安装配置

一、安装Python Agent
pip install apache-skywalking

二、配置并启动Agent
系统起始位置添加如下代码：
from skywalking import agent, config
#配置SkyWalking
config.init(agent_collector_backend_services=‘192.168.100.201:11800’, agent_instance_name=‘XZFLottery’)

#启动SkyWalking Agent
agent.start()

三、如果报错，检查变量名称
/usr/local/lib/python3.8/site-packages/skywalking/agent.py

def init(**kwargs) -> None:
“”"
Used to initialize the configuration of the SkyWalking Python Agent.
Refer to the official online documentation
https://skywalking.apache.org/docs/skywalking-python/next/en/setup/configuration/
for more information on the configuration options.

Args:
    **kwargs: Any of the configuration options listed
"""
glob = globals()

print(kwargs.items())

for key, val in kwargs.items():

    print(key)
    print(val)
    print(options)

    if key not in options:
        raise KeyError(f'Invalid configuration option {key}')

    glob[key] = val

四、官网
Legacy Setup
You can always fall back to our traditional way of integration as introduced below, which is by importing SkyWalking into your project and starting the agent.

Defaults
By default, SkyWalking Python agent uses gRPC protocol to report data to SkyWalking backend, in SkyWalking backend, the port of gRPC protocol is 11800, and the port of HTTP protocol is 12800,

See all default configuration values in the Configuration Vocabulary

You could configure agent_collector_backend_services (or environment variable SW_AGENT_COLLECTOR_BACKEND_SERVICES) and set agent_protocol (or environment variable SW_AGENT_PROTOCOL to one of gprc, http or kafka according to the protocol you would like to use.

Report data via gRPC protocol (Default)
For example, if you want to use gRPC protocol to report data, configure agent_collector_backend_services (or environment variable SW_AGENT_COLLECTOR_BACKEND_SERVICES) to :11800, such as 127.0.0.1:11800:

from skywalking import agent, config

config.init(agent_collector_backend_services=‘127.0.0.1:11800’, agent_name=‘your awesome service’, agent_instance_name=‘your-instance-name or ’)

agent.start()
Report data via HTTP protocol
However, if you want to use HTTP protocol to report data, configure agent_collector_backend_services (or environment variable SW_AGENT_COLLECTOR_BACKEND_SERVICES) to :12800, such as 127.0.0.1:12800, further set agent_protocol (or environment variable SW_AGENT_PROTOCOL to http):

Remember you should install skywalking-python with extra requires http, pip install "apache-skywalking[http].

from skywalking import agent, config

config.init(agent_collector_backend_services=‘127.0.0.1:12800’, agent_name=‘your awesome service’, agent_protocol=‘http’, agent_instance_name=‘your-instance-name or ’)

agent.start()
Report data via Kafka protocol
Please make sure OAP is consuming the same Kafka topic as your agent produces to, kafka_namespace must match OAP side configuration plugin.kafka.namespace

Finally, if you want to use Kafka protocol to report data, configure kafka_bootstrap_servers (or environment variable SW_KAFKA_BOOTSTRAP_SERVERS) to kafka-brokers, such as 127.0.0.1:9200, further set agent_protocol (or environment variable SW_AGENT_PROTOCOL to kafka):

Remember you should install skywalking-python with extra requires kafka, pip install “apache-skywalking[kafka]”.

from skywalking import agent, config

config.init(kafka_bootstrap_servers=‘127.0.0.1:9200’, agent_name=‘your awesome service’, agent_protocol=‘kafka’, agent_instance_name=‘your-instance-name or ’)

agent.start()
Alternatively, you can also pass the configurations via environment variables (such as SW_AGENT_NAME, SW_AGENT_COLLECTOR_BACKEND_SERVICES, etc.) so that you don’t need to call config.init.

All supported environment variables can be found in the Environment Variables List.