Supervision: manage the crisis, not suffer it anymore

What to do when the data center hosting the supervision that oversees it, suffers an incomplete or total failure? Multiple logics can be implemented. First of all, considering a SaaS supervision service can be the solution: the monitoring is carried out as close as possible to the elements to be supervised, but the restitution and the follow-up are carried out on an application, web or mobile, external to the Information System, and for which an availability SLA is operational, with specialized actors mobilized to reach it. The continuation of this post will thus treat the platforms of supervision on premise.

Traditionally, implementations are based on replication/synchronization of virtual machines providing the service by taking advantage of technologies inherent in any virtualization or hyper-convergence solution. Despite all the precautions taken, and the technological prowess in this field, a technical or operational problem can lead to unavailability of supervision.

Customers may also request high application availability across two data centers. This second approach involves a specific software architecture on all functional layers: user access, bidirectional communication between supervisors and the central platform, data storage, analysis. Above all, it implies very strong prerequisites on the data centers: dedicated inter-site link constantly available, network latency between sites less than 5ms.

The ServiceNav product offers such an architecture.

These approaches are costly, complex, and in fact rarely implemented.

Let's be provocative: in the event of an IT disaster, what will be the Data Center Manager's priority? Ensure operational continuity of supervision, or first maintain or restore application availability?

In the event of a total IT disaster, the first priority is to understand what is happening: supervision plays a crucial role here in contextualizing the failure encountered, and should therefore not be unavailable.

Try it free for 30 days

The ServiceNav architecture natively provides this visibility in the event of a complete or partial failure of the data center hosting the monitoring, impacting it.

The principle is the following, simple and very quick implementation:

1. Install a supervisor in the backup data center and connect it to the ServiceNav website - 15 minutes
2. Precisely define the set of controls needed to determine the status of the nominal data center :
   • Control points on all the components of the supervision
   • Control points on infrastructure components: network, storage, server, hypervision, etc.
   • The dependency relationships between these checkpoints
3. Activate the local web interface on this supervisor, allowing to visualize all the controls under the responsibility of this supervisor - 15 minutes

This way, in the event of an outage affecting all or part of the central monitoring website, you retain visibility into essential controls.

In detail: here.

Would you like to know more? Please contact us here !