It's always a challenge to optimize the performance of an application to be highly available and without customer-impacting latencies. For critical application services like databases, if there is a degradation in its performance, then all other services using that database will have their performance degraded or impacted. So, comprehensive visibility across the entire stack becomes critical. In many cases, however, monitoring tools providing needed visibility may consume the very same processing resources used by the monitored application itself. And the question becomes, how to monitor a critical application such as a database consistently and reliably without the negative performance hit or the need for disruptive fixes such as rebooting your critical server?
Read more