Browse Prior Art Database

System and method to effectively monitor the services hosted on the cloud

IP.com Disclosure Number: IPCOM000249513D
Publication Date: 2017-Mar-02
Document File: 4 page(s) / 79K

Publishing Venue

The IP.com Prior Art Database

Abstract

Cloud Monitoring System describes the effective and efficient way to Monitor the Cloud Services. If monitoring is not done in an efficient fashion, it can become a overkill. This article explains how the system can be monitored for availability without affecting the cloud services performance.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 40% of the total text.

1

System and method to effectively monitor the services hosted on the cloud

In a Cloud System, there are multiple services exposed. These multi-tenant services are expected to be up 24/7. So it’s important to monitor them for any failure. Any failure even for minutes would affect the Cloud System's ROI. It directly affects multiple customers who are using the services. Most of these services include database containers, data cache, IOT related etc.,

Monitoring the cloud system, means in general to identify the heartbeat of the systems and services. Though there may be many ways to monitor these and identify if they are up and running, one of mostly and heavily used is to send a REST call to the services. During peak hours, this being multi-tenant services, would receive the monitoring calls from the customers consuming these services and the provider of the cloud system. This problem is more than what we can think of as its three-fold. Individual services consumed by each customer will get the monitoring calls. This means that customer using more than one services, which is generally the case, would make more and more calls to monitor. And this means its multiple nodes of multiple services which are being monitored. And the host of cloud system also would host a web page for its customers to know what services are up and running at a given point in time.

During peak hours, load to the services for feature related calls would be more and top of it the calls for monitoring also would be critical. This loads the system, as the monitoring calls are made quite often for critical applications. This loads the service itself and also the dependent services which might end up saturated. For e.g. if the Mobile services are to be monitored, the dependent DB services also has to be monitored along with the mobile services. This proposal defines a methodology to make these monitoring calls to be effective and the aim is not to load the system with these calls

The core idea is to make effective and efficient monitoring calls to the services hosted in the cloud system. There can be multiple services with multiple nodes hosted in a cloud system. The proposal is to host a Cloud System Monitoring system, which monitors the services and the dependent services. Functionality of this Cloud Monitoring System(CMS) is to monitor the list of the services consumed by the registered service providers.

This CMS would monitor the sockets opened to these services. And if the socket stream is not responding, it would respond to the provider of the service or the dashboard where these services are being monitored by the cloud system provider. Socket Listener would be responsible for monitoring the services and alerting the providers.

With this proposal, the CMS system maintains a socket connection to a service for e.g. Cloudant database. It will also add a SocketClosedListener to it. This listener gets to know when the socket is closed or not responding. This means that the...