Cloud Monitoring

Cloud Monitoring Application Consists of two main scripts Reorter[reporter.py] and Manager[manage.py]

• This application can be use to monitor different instances running on cloud and send atomated emails if something went wrong on any instance

Reporter

• Reporter is the script that is run on each instance and collect data
• Data collected by reporter
  • CPUUtilization
  • DiskActivity
  • MemmoryUtilization
  • DiskUsage
  • NetworkActivity
  • MetaData

Reporter keep sending data to Monitor through a post request which can be use to monitor instance

The data collected by reported is exposed by endpoints:

• Available Endpoints
  • GET
    • /get/metrics : Returns all metrics data

    {
            "CPUUtilization": {
                "CPUUtilization": {
                    "Unit": "percentage",
                    "Value": "1.96"
                },
                "Time": "2020-06-16 19:32:16"
            },
            "DiskActivity": {
                "BytesRead/s": "3.09",
                "BytesWrite/s": "11.03",
                "ReadTransections/s": "0.41",
                "Time": "2020-06-16 19:32:16",
                "TotalTransections/s": "3.50",
                "WriteTransections/s": "0.41"
            },
            "Diskutilization": {
                "Avail": "6.6G",
                "Hostname": "ip-172-31-43-219.ap-south-1.compute.internal",
                "MountPoint": "/",
                "Percentageused": "19",
                "Time": "2020-06-16 19:32:16",
                "Totalsize": "8.0G",
                "Used": "1.5G"
            },
            "MemmoryUtilization": {
                "Available_Mem": "597M",
                "Mem_used": "231M",
                "MemoryUtilized": {
                    "Unit": "percentage",
                    "Value": "39"
                },
                "Time": "2020-06-16 19:32:16",
                "Total_Mem": "983M"
            },
            "Metadata": {
                "ami-id": "ami-0447a12f28fddb066",
                "hostname": "ip-172-31-43-219.ap-south-1.compute.internal",
                "instance-id": "i-08e80892f2cb506cd",
                "instance-type": "t2.micro",
                "public-ipv4": "13.126.53.116"
            },
            "NetworkActivity": {
                "Time": "2020-06-16 19:32:16",
                "rxkB/s": "32.85",
                "rxpck/s": "28.54",
                "txkB/s": "4.28",
                "txpck/s": "10.98"
            },
            "Time": "2020-06-16 19:32:16"
        }
    

    • /get/graph Can be use to make graph of a perticular metrics Query parameter to pass

      • interval = seconds
      • metric = name of metric

    • /get/metadata Returns all metadata about the instance

  • /update/status Can be use to stop or resume the reporter app Query parameter to pass
    • status = {pause,resume,stop}
  • /update/interval Update the time after which the the reporter sends post request to manager

Testing Instance

By default, the Replorter app will use port 5000,

http://13.126.53.116:5000/get/metadata

Manager

• Manager will recives post request from all the reporter running on different instance
• Maintain record of every instance in a databse
• Checks different metrics and sends notification to the user in form of email or slack message.
  • CPUCheck
  • MemoryCheck
  • DiskCheck
• Available Endpoints
  • POST
    • /app/post/data Data is sent to this endpoint in json format