Enabling Cloud-Ready Monitoring
Is Your Monitoring System Cloud Ready?
Cloud computing offers the advantages of pay-per-use pricing and the ability to make computing resources available quickly when and where they are needed. As organizations begin to deploy business services using cloud computing resources, understanding how to monitor and manage these cloud-based services becomes critical.
Below are some key considerations in determining if your monitoring system is cloud-ready.
Deployment agility :
Just as the cloud's underlying infrastructure is agile, the monitoring system needs to be nimble and agile as well, to enable you to quickly turn on applications in the cloud, manage them and start delivering business-critical services. Also, deployment should be automated as much as possible to reduce human intervention, and should be a simple process taking a few minutes, not days or weeks.
Ubiquitous access :
The monitoring system should be web-based so it is accessible from anywhere at any time. This way, monitoring can be done from within the enterprise as well as beyond the enterprise.
Ability to operate in a secure manner, within and across firewalls :
Since a business service may include servers within the enterprise and other servers in the cloud, the monitoring system should have the capability to operate within enterprise networks and also across firewalls. Further, for high security, monitoring agents, if used, should not listen on any additional TCP ports. Where possible, all communication should be protected using industry standard encryption protocols. Many monitoring systems that use SNMP and other proprietary protocols do not meet this requirement.
Capability to monitor the performance of applications hosted in the cloud :
The monitoring system should be able to measure the performance of applications, whether they reside in the enterprise or in the cloud. The metrics collected can serve as a benchmark for determining if there is a problem and what is causing it; e.g., is it due to an application within the enterprise, or one hosted in the cloud?
Capability to monitor remote desktops efficiently :
Desktops as a service is one of the popular applications driving cloud computing. Desktop virtualization technologies allow the efficiency of dozens of desktops on a single physical server. A cloud monitoring solution should parallel that efficiency by allowing detailed monitoring inside each desktop without requiring a monitoring agent on each desktop.
Extensibility to support cloud provider APIs :
When a performance bottleneck is detected, knowing what caused the problem is the first step in fixing it. Is the problem caused by a malfunctioning application, or did performance degrade because usage of the application in the cloud has exceeded the paid-for-usage limit? Each cloud provider offers a custom interface for monitoring the performance and usage of the cloud. The monitoring system should be easily extensible to support any cloud provider API in order to help isolate the root cause of performance issues.
End-to-end monitoring, with automated root cause analysis :
Ultimately, all business users care about is whether their critical services are meeting customer expectations for service performance and, if not, why. This capability, it is essential for monitoring end-to-end across a multi-tiered infrastructure that can span public cloud, private cloud and intra-enterprise networks. The ability to monitor what is happening at every layer of every component in such an end-to-end IT infrastructure- and automatically isolate which layer of which component is the source of an anomaly-- goes to the purpose of the Event Management process as defined by the IT Infrastructure Library (ITIL): "to detect events, makes sense of them, and determine the appropriate control action."
Ability to deliver monitoring itself as a cloud-based service :
The monitoring tool itself should be delivered as a cloud-based service, so it can be turned on and off as the business demands. Additionally, the cloud monitoring tool should be provisioned and decommissioned on a pay-per-use model.
Licensing flexibility :
Cloud computing environments can be scaled as demand increases or decreases. The cloud monitoring system must be flexible in handling this. Furthermore, since the server sizing can also vary based on load, monitoring solutions that are licensed per physical or virtual CPU, core, socket, memory or the operating system being used are not practical for cloud environments. Look for a solution that is not licensed based on the underlying infrastructure but rather based on need and usage.
Monitoring Cloud Infrastructures Using eG Enterprise
Legacy monitoring solutions and frameworks struggle to deal with the new requirements that cloud infrastructures pose. These solutions are complex to install and configure and they have rigid licensing policies that are based on applications monitored and the number of CPUs, cores, sockets, etc. Being based on SNMP or other proprietary protocols, these solutions cannot handle the communication requirements that cloud-based services have.
Using the key considerations discussed above, the table below summarizes how eG Enterprise addresses all of the Cloud Ready monitoring needs:
| Cloud-Ready Monitoring Requirements
||eG Enterprise Capabilities
|| eG Enterprise agents can be deployed in minutes. A silent installation procedure can be used to install agents without any human intervention. The eG agent automatically discovers the major applications running on a server and begins detailed monitoring and reporting for each application. Furthermore, alerting thresholds are set automatically and vary dynamically to account for normal patterns, which reduces false alerts.
| Ubiquitous access
|| A 100% web-based interface ensures that users can log-in from anywhere and monitor the health of their critical business services and its underlying infrastructure - whether based on a private cloud, a public cloud, or conventional enterprise infrastructure.
| Ability to operate within and beyond firewalls
|| eG Enterprises uses a 100% web-based architecture. All communication between the agents and the manager are over HTTP/HTTPS. This means no firewall rules need to be changed or ports opened to allow management traffic to flow. The agents also do not listen on any TCP ports.
| Capability to monitor performance of applications hosted in the cloud
|| eG Enterprise includes the ability to monitor applications using both active and passive approaches. Out of the box, eG Enterprise supports over 120+ applications, including most popular web, database, middleware, messaging and other applications.
| Capability to monitor remote desktops efficiently
|| eG Enterprise's patented In-n-Out MonitoringTM technology tracks the top CPU- and memory-consuming processes in each virtual desktop without requiring an agent on the desktop. It is also possible to monitor both VDI and Citrix or Terminal Server users from the same single pane of glass
| Extensibility to monitor cloud provider APIs
|| The eG Integration Console can be used to extend eG Enterprise to monitor different cloud providers.
| End-to-End monitoring with automated root cause analysis
|| eG Enterprise was designed from the ground up to monitor multiple components across multiple domains, and automatically distinguish the cause of service-affecting problems from their effects. With its end-to-end view, eG Enterprise can rapidly pinpoint the root cause of response time problems from all angles, such as application malfunction, unusually high workload, the cloud provider not providing the requested resources, or improper provisioning of a cloud-hosted server. It automatically downgrades the severity of alerts that are not root cause issues, thus drastically reducing troubleshooting time.
| Deliver monitoring as a cloud-based service
|| eG Enterprise is available as a SaaS offering from eG Innovations and from other service provider partners. Customers can choose to turn monitoring on and off as their business needs dictate.
| Licensing flexibility
|| eG Enterprise's unique single agent licensing model offers unparalleled flexibility in deployment. A single agent license can be used to perform in depth layer model monitoring of one of over 120 different applications and the underlying VM and server it is running on regardless of the server's hardware capabilities, or the OSs running on it. A single agent can also be used to remotely monitor multiple VDs on multiple servers, or multiple network and storage devices.