Frequently Asked Questions

About the StackState technology

There are multiple sources to enrich the topology of your environment, but the main categories are:

  • StackState Agent; automatically discovers all processes connecting with each other via network connections. The StackState Agent sends a report of these processes + connections every N seconds to StackState (N is configurable).
  • Provisioning and deployment tools like Chef, Puppet, Ansible and XL Deploy
  • Service discovery systems, like Consul and Evolven
  • Orchestration tools, like Kubernetes and Mesos
  • Existing tools and/or systems that already know the topology, like HP uCMDB and ServiceNOW
  • Tracing tools

StackState is capable of retrieving information about components and their relations via any source and can merge topologies into one single model.

From this information StackState can infer that some processes are running on a specific host. With this information it can create a component for the process and the host and it can create a dependency from the process component to the host component. Further it can infer (based on the network connection information that contains both the local and remote ip address + port) that there is a remote component involved. This information is used to create the remote component and a dependency between both the local and the remote component.

StackState uses a mapping in combination with a template to determine how to create a component based on the process information. For example a MySQL process will be recognized and mapped to the MySQL template (that StackState provides). A similar process that has a “java” command line will be mapped to the Java template. There are defaults for these, however this can be customized to map to different templates.

StackState is also able to follow business transactions through the different services to discover the business steps in business processes.

StackState is capable of maintaining an automated synchronization of all the components and relations. It will automatically discover new-, updated- and/or deleted components.
StackState is an open platform that can support any technology ea. Splunk, Elasticsearch, Zabbix, Graphite, OpsView, Slack, HP UCMDB, ServiceNow and AppDynamics. In addition the StackState agent supports currently 80+ technologies. For a detailed list please see our integration page on the website. Lastly, StackState has a developer connection (REST API) available.
There are various ways to integrate with StackState:

  • The StackState Agent supports plugins that can be used to extend and pull events and/or metrics from your tool and push them to StackState.
  • Add your own code to the agent using Python. This way you can just call functions like ‘send_event’ to send an event or ‘gauge’ to send metrics.
  • Use the StackState’s REST API directly to push metrics/events/topology and other telemetry information to StackState.
  • Create a plugin for StackState using our open plugin API to support the specific tool.
StackState is designed to support massive scale of components (business processes, servers, VM’s, etc). Current tests have shown StackState can scale up to a million+ components.
APM tools are focused on IT components, metrics and events with regards to applications and showing that in a display. StackState has an all-embracing scope. StackState focuses on your business processes and the supporting IT components with the dependencies. StackState will create a model of the entire IT stack from the hardware level all the way up to the business process level. As such StackState gives a total overview of your whole enterprise and not just the IT part.
StackState also leverages real-time metrics, events and traces to proactively monitor your business. StackState gives a holistic view of your business, allowing slicing and dicing this into views for the different departments in your organization (from business process owner to the DevOps engineer and/or datacenter operator).
StackState’s model integrates information from provisioning and log/event management tools. StackState provides automated root cause analysis. This will show all affected components in the chain and/or service for all different teams and tools.
We support role-based views. You can define whichever selection from the overall model (slice and dice).
Yes, if metrics and topology information from the existing infrastructure can be pushed to StackState. If that is not possible, we could use a separate tool (possibly managed by StackState) that pulls data from custom sources and pushes them into StackState.
There are various ways to send data to StackState:

  • The StackState Agent supports plugins that can be used to extend and pull events and/or metrics from your tool and push them to StackState.
  • Add your own code to the agent using Python. This way you can just call functions like ‘send_event’ to send an event or ‘gauge’ to send metrics.
  • Use the StackState’s REST API directly to push metrics/events/topology and more to StackState.
  • Create a plugin for StackState using our open plugin API to support the specific tool.
The StackState model supports containers and micro services.

Can’t find your answer?

Get in touch