VMware vCenter Integration

This page describes the resource requirements, compatibility with different Docker®-based software, and the specifications required to integrate VMware vCenter with Cisco Observability Platform.

VMware vSphere is a virtualization platform that transforms data centers into aggregated computing infrastructures that include CPU, storage, and networking resources. vCenter Server is a core component of VMware vSphere that allows you manage multiple hosts connected in a network and pool host resources. You can integrate VMware vCenter with  Cisco Observability Platform to view the metrics data on the  Cisco Observability Platform UI.

The vCenter collector collects the VMware monitoring metrics for cluster, network, host, datastore, resource_pool, storage and Virtual Machines (VM).

To monitor the VMware data, you require:

• vCenter collector: Install this collector in the required virtual machine to collect the data from that machine to the  Cisco Observability Platform.
• aiopscore solution: When you subscribe to AIOps module, this solution is added by default. This solution includes the files with all the configurations that determine what all entities and their correlation gets displayed on the UI.

This document contains references to the Docker documentation. Cisco does not own any rights and assumes no responsibility for the accuracy or completeness of such third-party documentation.

Before You Begin

Before installing vCenter collector for data ingestion, ensure that you meet the following requirements:

• You are connected to vCenter from the server. Run the following command to verify:
  

  $sudo telnet vcenterserver 

  CODE
• You have sudo privileges on the server to run the Docker commands

• You have vCenter collector authentication credentials

  

  vCenter collector is required to export the data from collector to the  Cisco Observability Platform. The collector requires client_id, client_secret and token_url. Contact your  Cisco Observability Platform administrator for the OAuth credentials.

• (Conditional) To use FSOC queries to validate the data ingestion, ensure that you:

  1. install FSOC on your device by using the instructions to Install and Configure FSOC.
  2. configure the profile to FSO API using auth credentials

Software Requirements

The vCenter collector can run in hybrid (Linux and Windows) or Linux-only clusters. The vCenter collector requires:

• VMware >= v6.5
• Docker >= 18.09.2
• Docker-compose with sufficient licenses
• Access to the following components:

Install vCenter Collector

The Cisco Observability Platform uses the OpenTelemetry Protocol as its native MELT telemetry data format. It uses the Common Ingestion Services to provide the endpoints for the OTLP telemetry data ingestion. Perform the following steps to install vCenter collector on a Docker container:

1. Ensure that your environment meets the Software Requirements.
2. Log in to the server where you require to install the collector.
3. Setup Docker and Docker compose on the server. See, Docker and Docker Compose.

4. Validate the Docker and Docker Compose versions.     

   Validate-Example

   $sudo docker -v

   CODE

   $sudo docker compose version

   CODE

   
   Following is an example output:

   Example-output

   $sudo docker -v
   Docker version 23.0.1, build a5ee5b1
   $sudo docker compose version
   Docker Compose version v2.16.0

   CODE

5. Open JFrog and go to the location maven-releases>com>appdynamics>aiops>aiops-agt-vmware to download the latest files of vCenter collector.
   
   

6.  On the collector-config-vcenter.yaml file, update the following details. This is the yaml file where all the mandatory configuration must be updated. The yaml file contains vCenter and the platform properties. You must update the required values to connect to vCenter and send metrics to vCenter collector API.

   1. Update vCenter connection details in the configuration.

      vcenter:
          endpoint: <endpoint>
          username: <user>
          password: <password>
          collection_interval: 1m
          tls:
            insecure: true

      CODE

   2. Update the vCenter collector API endpoint for sending metrics and events to Cisco Observability Platform.

      exporters:
        otlphttp/withauth:
          endpoint: https://<tenant_name>.observe.appdynamics.com/data
          auth:
            authenticator: oauth2client

      CODE

   3. Update the Cisco Observability Platform oauth credentials in extensions

      extensions:
        oauth2client:
          client_id: "***************"
          client_secret: "*******************"
          token_url: "https://<tenant_name>.observe.appdynamics.com/auth/********************/default/oauth2/token"
          timeout: 60s

      CODE
      

      You can update the timeout value. Consider the timeout setting based on the number of applications and resources running on the VMware cluster.

7. Run the docker compose command to start the container and send metrics to Cisco Observability Platform.

   $sudo docker compose up -d

   CODE

8. Validate the Container status.

   $ docker ps

   CODE

   The output is similar to the following:

   CONTAINER ID   IMAGE                                         COMMAND                  CREATED       STATUS       PORTS                                                                                                             NAMES
   cf76da77999c   otel/opentelemetry-collector-contrib:0.88.0   "/otelcol-contrib --…"   11 days ago   Up 11 days   0.0.0.0:5556->5556/tcp, :::5556->5556/tcp, 4317/tcp, 55678-55679/tcp, 0.0.0.0:6667->6667/tcp, :::6667->6667/tcp   vcenter-collector

   CODE

Validate vCenter Data

After installing the vCenter collector, you need to validate the metrics and events that are ingested into the Cisco Observability Platform.  You can validate the data by using either Query Builder or fsoc.

Data Validation using Query Builder 

You can validate the data by using Query Builder. Open Query Builder and enter the following query:

fetch events(aiopscore:vmware_event) { attributes} order events.desc() limits events.count(50) since -1h until now()

Data Validation using FSO CLI (FSOC) 

You can validate the data by using FSOC. To use FSOC, see Install and Configure FSOC and configure the profile to vCenter collector API using authentication credentials.

Perform the following to validate if the latest data is getting included through the collector:

1. Setup the FSOC profile using the service principle JSON file (secret file) provided by your company administrator. 
   

   $sudo fsoc config set --profile <name-of-your-profile> --secret-file <name-of-secret-file>

   CODE

   For example,

   $sudo fsoc config set --profile demo --secret-file ciscoitdemo-access.json

   CODE

2. Use the profile using fsoc config command.
   

   $sudo fsoc config use --profile <name-of-your-profile>

   CODE

   For example,

   $sudo fsoc config use --profile demo 

   CODE

    

3. Run UQL queries after successfully setting up FSOC profile.
   For information about UQL queries, see the Unified Query Language User Guide.
   

   This is an example query and output:

   $sudo <user> resources % fsoc uql "fetch events(aiopscore:vmware_event) { attributes} order events.desc() limits events.count(50) since -1h until now()" 
   ✓ Platform API call (POST /monitoring/v1/query/execute)
    events                                                                                        
    attributes                                                                                    
    name                               | value                                                    
   ================================================================================================
    aiopscore.vmware_event.alarm_name  | AiOps VM CPU Usage - V1                                  
    appd.event.type                    | aiopscore:vmware_event                                   
    aiopscore.vm.type                  | vm                                                       
    description                        | raw message                                              
    appd.isevent                       | true                                                     
    aiopscore.vmware_event.severity    | Normal                                                   
    reason                             | raw text                                                 
    aiopscore.vmware_event.object      | aiops-lab-dev1                                           
    severity                           | Normal                                                   
    _parsing_status                    | false                                                    
    aiopscore.vm.hostname              | aiops-lab-dev1                                           
    aiopscore.display_name             | AiOps VM CPU Usage - V1                                  
    parsing_error_or_warnings          | Missing mandatory parsing attribute _message_parser.type.
    namespace                          | aiopscore                                                
    subtype                            | text                                                     
    telemetry.sdk.name                 | aiops-fsoc-exporter                                      
    aiopscore.vmware_event.object_type | Host                                                     
   ------------------------------------+-----------------------------------------------------------
    aiopscore.vmware_event.alarm_name  | AiOps VM CPU Usage - V1                                  
    appd.event.type                    | aiopscore:vmware_event                                   
    aiopscore.vm.type                  | vm                                                       
    description                        | raw message                                              
    appd.isevent                       | true                                                     
    aiopscore.vmware_event.severity    | Critical                                                 
    reason                             | raw text                                                 
    aiopscore.vmware_event.object      | aiops-lab-dev1                                           
    severity                           | Critical                                                 
    _parsing_status                    | false                                                    
    aiopscore.vm.hostname              | aiops-lab-dev1                                           
    aiopscore.display_name             | AiOps VM CPU Usage - V1                                  
    parsing_error_or_warnings          | Missing mandatory parsing attribute _message_parser.type.
    namespace                          | aiopscore                                                
    subtype                            | text                                                     
    telemetry.sdk.name                 | aiops-fsoc-exporter                                      
    aiopscore.vmware_event.object_type | Host                                                     
   ------------------------------------+-----------------------------------------------------------

   CODE

Data Ingestion

In VMware, you can use environment variables in a script and attach them to multiple alarms. See VMware Documentation.

The following table provides the mapping between the environment variables available in Cisco Observability Platform and VMware that you can use in your script:

Script

Use the following script to send alarm events from VMware to Cisco Observability Platform:

#!/usr/bin/env python
 
import requests
import json
import os,sys
import getpass
 
print("myuser",getpass.getuser())
#sys.stderr = open('/var/local/vc_errors.txt','a')
 
def get_access_token(url, client_id, client_secret):
    response = requests.post(
        url,
        data={
            "grant_type": "client_credentials",
            "client_id": client_id,
            "client_secret":client_secret
        }
    )
    print("auth_token response: ", response)
    return response.json()["access_token"]
 
def generate_payload():
    output=[]
    data={}
    data["route"] = "vm"
    data['aiopscore.vmware_event.alarm_id'] = os.getenv('VMWARE_ALARM_ID')
    data["aiopscore.vmware_event.alarm_name"] = os.getenv('VMWARE_ALARM_NAME')
    data["aiopscore.vmware_event.object"] = os.getenv('VMWARE_ALARM_TARGET_NAME')
    data["aiopscore.vmware_event.object_type"] = "VM"
    data["aiopscore.vmware_event.severity"] = os.getenv('VMWARE_ALARM_NEWSTATUS')
    data["aiopscore.vmware_event.triggered_time"] = ""
    data["aiopscore.vmware_event.acknowledged_time"] = ""
    data["aiopscore.vmware_event.acknowledged_by"] = ""
    data["aiopscore.vmware_event.event_description"] = os.getenv('VMWARE_ALARM_EVENTDESCRIPTION')
    data["aiopscore.vmware_event.alarm_value"] = os.getenv('VMWARE_ALARM_ALARMVALUE')
    data['aiopscore.vmware_event.event_compute_resource'] = os.getenv('VMWARE_ALARM_EVENT_COMPUTERESOURCE')
    data['aiopscore.vmware_event.triggering_summary'] = os.getenv('VMWARE_ALARM_TRIGGERINGSUMMARY')
    data["aiopscore.vmware_event.target_id"] = os.getenv('VMWARE_ALARM_TARGET_ID')
    data["aiopscore.vmware_event.alarm_old_status"] = os.getenv('VMWARE_ALARM_OLDSTATUS')
    data['aiopscore.vmware_event.alarm_new_status'] = os.getenv('VMWARE_ALARM_NEWSTATUS')
    data['aiopscore.vmware_event.declaring_summary'] = os.getenv('VMWARE_ALARM_DECLARINGSUMMARY')
    data['aiopscore.vmware_event.event_username'] = os.getenv('VMWARE_ALARM_EVENT_USERNAME')
    data['aiopscore.vmware_event.event_host'] = os.getenv('VMWARE_ALARM_EVENT_HOST')
    data['aiopscore.vmware_event.event_vm'] = os.getenv('VMWARE_ALARM_EVENT_VM')
    data['aiopscore.vmware_event.event_network'] = os.getenv('VMWARE_ALARM_EVENT_NETWORK')
    data['aiopscore.vmware_event.event_datastore'] = os.getenv('VMWARE_ALARM_EVENT_DATASTORE')
    data['aiopscore.vmware_event.event_dvs'] = os.getenv('VMWARE_ALARM_EVENT_DVS')
    output.append(data)
    return output
 
def publish_events():
    endpoint = os.getenv("endpoint")
    tenant = os.getenv("tenant")
    auth_url = "https://{}/auth/{}/default/oauth2/token".format(endpoint, tenant)
    url = "https://{}/rest/aiopsservices/webhook/api/v1/events".format(endpoint)
    client_id = os.getenv("client_id")
    client_secret = os.getenv("client_secret")
    token = get_access_token(auth_url, client_id, client_secret)
    headers = {
      'Content-Type': 'application/json',
      'Authorization': 'Bearer {}'.format(token)
    }
    print(headers)
    data=generate_payload()
    print(json.dumps(data))
    response = requests.post(url, data=json.dumps(data), headers=headers)
    print(response.text)
 
#publish_events()
 
if __name__ == "__main__":
   publish_events()

Next Steps

• • Monitor the VMware entities.

Third party names, logos, marks, and general references used in these materials are the property of their respective owners or their affiliates in the United States and/or other countries. Inclusion of such references are for informational purposes only and are not intended to promote or otherwise suggest a relationship between Cisco AppDynamics and the third party.