What Is EVA?

Customized Python Script

def condition():
    e1 = eva.Event()
    kpi1 = e1.addkpi()
    e1.ret = eva.avg()     
    s1 = eva.Strategy()
    s1.formula = e1
    action1 = eva.log()
    s1.addaction(action1)

In the preceding script:

• Event: describes data subscription, processing, and judgment.
• LogEvent: describes the subscription to device logs.
• Strategy: describes the orchestration and execution actions of events.
• Action: defines device behaviors, including command delivery and data calculation.

Customized JSON Script

• An example of a single-task script is as follows:

  {
  	"ItemName": {
  		"description": "",
  		"define": {},
  		"events": {},
  		"strategy": "",
  		"tasks": {}
          }
  }

• An example of a PMI script is as follows:

  {"ItemName1": {"description": "","define": {},"events": {},"strategy": "","tasks": {}}}
  {"ItemName2": {"description": "","define": {},"events": {},"strategy": "","tasks": {}}}
  {"ItemName3": {"description": "","define": {},"events": {},"strategy": "","tasks": {}}}

In the preceding scripts:

• description: provides brief description of the script.
• define: defines global variables for the entire script. A maximum of 15 global variables can be defined. Other variables cannot be referenced.
• event: defines the entry conditions for running a script.
• strategy: defines an event or an event combination condition for executing a task or an action.
• task: defines an action triggered by a strategy and the relationship between actions. Each task contains one action. If the service logic is complex, you can define multiple tasks and define in a task the jump relationship between the task and other tasks. The service logic of the loop type is also defined in a task.

Configuring EVA Based on JSON PMI Scripts

In an inspection scenario, a customized JSON script is stored on an SFTP server. O&M personnel want to use EVA to collect routing information, interface diagnostic information, and device health information from devices. The JSON script is an inspection script that defines the commands for collecting routing information, interface diagnostic information, and device health information.

Networking diagram of performing inspection on devices through EVA

The configuration roadmap is as follows:

1. Compile a JSON script and define events, strategies, and tasks in the script.

2. Upload the JSON script to the device.

3. Install and register the JSON script.

Step 1 Compile a JSON script and define events, strategies, and tasks in the script.

{"inspection_1":{"description":"Collect IP routing information","events":{"e1":{"trigger":"eva.singleCollect()"}},"strategy":"e1","tasks":{"mainTask":{"parameters":{"IProute_cmdArray":["display bgp peer","display isis peer","display ospf peer"]},"action":"eva.cliArray(\"user\",${IProute_cmdArray})"}}}}  # Collect routing information.
{"inspection_2":{"description":"Collect statistics on error packets and diagnosis information on the interfaces","events":{"e1":{"trigger":"eva.singleCollect()"}},"strategy":"e1","tasks":{"mainTask":{"parameters":{"Interface_cmdArray":["display interface troubleshooting","display interface counters errors","display interface brief"]},"action":"eva.cliArray(\"user\",${Interface_cmdArray})"}}}}  # Collect interface diagnostic information.
{"inspection_3":{"description":"Collect system management information","events":{"e1":{"trigger":"eva.singleCollect()"}},"strategy":"e1","tasks":{"mainTask":{"parameters":{"Health_cmdArray":["display debugging","display device","display health"]},"action":"eva.cliArray(\"user\",${Health_cmdArray})"}}}}  # Collect device health information.

Step 2 Upload the JSON script to the device.

[~DeviceA] sftp client-transfile get host-ip 10.2.1.1 username client001 password Helloworld@6789 sourcefile CollectInformation.json
# Configure DeviceA functioning as an SFTP client to download the JSON script CollectInformation.json from the SFTP server.

Step 3 Install and register the JSON script.

<DeviceA> install eva script CollectInformation.json inspection

Configuring EVA Based on Python Scripts

A customized Python script is stored on an SFTP server. O&M personnel want to use EVA to monitor the CPU usage and memory usage. The Python script specifies that when the CPU usage and memory usage exceed 90%, the system automatically generates an alarm log reporting high CPU and memory usages, prompting O&M personnel to maintain devices in a timely manner.

Networking diagram of monitoring CPU usage changes through EVA

The configuration roadmap is as follows:

1. Compile a Python script to customize events and troubleshooting strategies.

2. Upload the Python script to the device.

3. Install and register the Python script.

Step 1 Compile a Python script to customize events and troubleshooting strategies.

import eva
def condition(): 
    e1 = eva.Event()
    kpi1 = e1.addkpi("huawei-cpu-memory:cpu-memory/board-cpu-infos/board-cpu-info/system-cpu-usage")
    e1.ret = eva.avg(kpi1, 10) > 90
 
    e2 = eva.Event()
    kpi2 = e2.addkpi("huawei-cpu-memory:cpu-memory/board-memory-infos/board-memory-info/os-memory-usage")
    e2.ret = eva.avg(kpi2, 10) > 90
    
    s1 = eva.Strategy()
    s1.formula = e1 & e2
    s1.validTime= 10
    action1 = eva.log("board ${e1.slot-id}--${e1.cpu-id} cpu and memory simultaneously overload")
    s1.addaction(action1)
# Compile the script cpuMemHigh.py to subscribe to the system CPU usage and memory usage, and calculate the average values of last 10 CPU usages and memory usages. If both the average values exceed 90%, an alarm log is generated.

Step 2 Upload the Python script to the device.

[~DeviceA] sftp client-transfile get host-ip 10.2.1.1 username client001 password Helloworld@6789 sourcefile cpuMemHigh.py
# Configure DeviceA functioning as an SFTP client to download the Python script file cpuMemHigh.py from the SFTP server.

<DeviceA> ops install file cpuMemHigh.py                           # Install the Python script.
<DeviceA> ops run python evamain.py install cpuMemHigh.py        # Register the Python script.