Chronosphere metrics have a variety of limits, depending on where your data is in the
processing cycle.
Your license limits the amount of data you
can write to and store within Chronosphere.
Metrics quotas assign specific percentages of your total
persisted writes limit to pools of metrics. Use quotas to limit groups of metrics and
ensure Chronosphere retains the highest value metrics.
See Managing invalid metrics
for information about debugging invalid metrics.
Size limits
Metrics have the following size limits, which help ensure consistency and prevent
issues with queries:
- Label names: 512 bytes
- Label values: 1,024 bytes
- Max number of labels: 64 labels
- Total time series bytes: 4,096 bytes
If label values exceed 1,024 bytes, Observability Platform shortens the label values
to accept the time series.
Example metric with a long label value
Example metric with a shortened value
Late-arriving metrics
Prometheus, OpenTelemetry, and other metric formats include one or more timestamp
values for each data point. These values indicate when a sample was observed, or
represent the time range of the data point. Chronosphere Observability Platform can
accept late-arriving data points within a time frame, depending whether the data
point matches an aggregation rule or is ingested without aggregation:
- Aggregation rules accept data points from two minutes to eight minutes past the
current ingestion time.
- Raw data points can be written to the database up to two hours before the ingestion
timestamp.
Future-arriving metrics
Observability Platform accepts incoming metrics one minute ahead of the current
timestamp. If metrics arrive outside of this one-minute buffer, Observability
Platform rejects those metrics and returns an error indicating that metrics arrived
too far into the future. 
