Monitoring and analytics in the SDDC
As discussed at the beginning of this chapter, the SDDC introduces some challenges, which cannot be easily overcome with traditional monitoring systems. This becomes clear if one looks at the traditional versus the SDDC way of deploying services and workloads.
In the traditional data center, workloads are often deployed in form of projects. They have a distinct function (web server, application server, database, and so on) as well as foreseeable workload profile. Based on this, the monitoring admin can set a set of thresholds to make sure that the workload is working within its expected range. Normally, these thresholds are CPU usage, memory usage, swapping, disk space, and so on.
A monitoring system is aware of the new server and is associating all these thresholds to the server. If one of these values are violated, it will send a warning or an alarm to the monitoring team or the administrator. This has been used for years in the data center and is a well...
