Operations database

An operations database ( OpsDB ) is a database that collects metrics of business processes as well as application and system statistics and makes them available for analysis purposes.

application

Use of an OpsDB

An operations database is used to collect data on the behavior of a business solution under certain conditions. The data can be used to fill a dashboard and to visualize the current status of the availability of business processes. It is also possible to evaluate the data in order to enable a prediction of future application behavior.

For example, consider a business process (e.g., a sales process) that depends on the availability of several systems. The OpsDB collects information on the status of the individual systems. If a system fails, errors occur in the systems that are dependent on the failed system and are noted in the OpsDB. The evaluation of the status of the system in the dashboard can indicate the failure of the business process (e.g. with a red traffic light).

Another possible application is the evaluation of the history, which makes it possible to correlate the failure of a business process with the status of the individual systems. This also enables a prediction of the behavior of the system under possible future conditions. This includes B. the behavior of the system under load, for example when a sales system has to reckon with a particularly high number of users due to a public holiday campaign. The prediction enables the planning of tests of the system in development as well as the (hardware) resources to be provided.

Comparison of monitoring technologies

technology	history	forecast	status	behavior
Logging
Monitoring
Operations database

structure

Structure of an OpsDB

An OpsDB maps a business process to be monitored that is dependent on one or more nodes (services and applications, servers , databases , web services , batch processes, etc.). These nodes can in turn be dependent on other nodes.

For each node observations (be Observation s ) collected that provide information about the state of the system.

For observations also can expectations ( Expectation s example, the duration of which require a process at most, an area that there is a measured value in the, or a state) are defined, in which an application is to be located.

Data That Is Collected

Operations databases collect a large number of parameters.

Overview of data in operations databases

category	Examples
Network traffic	Number and frequency of page requests Number of active and executed transactions Number of active and handled sessions
State of resources	Available and used resources Maximum resources used Number of additionally created or released resources Number of blocked threads waiting for a resource
Database connection	Number of database errors Number of database queries Average duration of database queries
Integration points and interfaces	State of fuses ( circuit breaker ) Number of requests Number of request timeouts Average response time to requests Number of positively answered requests Network errors and protocol errors Application error Addresses of the endpoints Number and maximum number of requests processed at the same time
Cache state	Number of items in the cache Memory usage Cache hit and cache miss rates Information about the garbage collector Maximum size of the cache Time to store objects in the cache
↑ Resources are elements such as CPU utilization, memory consumption and required (virtual) servers.

See also

• Operational Database (OLTP)

Web links

• Health monitoring. In: Microsoft Docs. Microsoft, July 1, 2019, accessed October 1, 2019 . 

credentials

1. ↑ ^{a b c d} Michael T. Nygard: Release It! Design and Deploy Production-Ready Software. O'Reilly, 2007, ISBN 978-0-9787392-1-8 , 17.7 Operations Database (English, 326 pages).