tlas.atlas-wiki.1.4.3.source-code.Metrics-and-Logging.md Maven / Gradle / Ivy
Atlas is intended for use with dimensional time-series data, however, one thing we often see is users trying to abuse it as a logging or event system. While there are some similarities, it is important to distinguish between these use-cases and use the more appropriate tool.
A time series is a set of values measured over time. An log message or event is represents a specific occurrence of something. This distinction is important for the types of queries and use-cases that can be supported efficiently by a given system.
## Uniqueness
Atlas is intended for showing time series that summarize events or behavior over time, not for finding or investigating a specific event.
## Granularity
## Log Parsing
It is a common use case to build log parsers to count events of a particular class (i.e. HTTP 2XX, HTTP 4XX) and send this data into Atlas. The key here is being able to observe the trend of events being reported to logs.
© 2015 - 2024 Weber Informatics LLC | Privacy Policy