What are the common PromQL functions and how to use them?

PromQL (Prometheus Query Language) is the query language for Prometheus, with common functions including:

Aggregation Functions:

• sum() - Summation
• avg() - Average
• max() - Maximum
• min() - Minimum
• count() - Count
• count_values() - Count occurrences of each value

Time Functions:

• rate() - Calculate average growth rate over a time window
• irate() - Calculate instantaneous growth rate (more sensitive)
• increase() - Calculate increment over a time window
• delta() - Calculate difference

Mathematical Functions:

• abs() - Absolute value
• ceil() - Round up
• floor() - Round down
• round() - Round to nearest
• exp() - Exponential
• ln() - Natural logarithm

Prediction Functions:

• predict_linear() - Predict future values based on linear regression

Example Queries:

promql
# Calculate QPS
rate(http_requests_total[5m])

# Calculate memory usage rate
sum(container_memory_usage_bytes) / sum(container_spec_memory_limit_bytes) * 100

# Predict when disk space will run out
predict_linear(node_filesystem_avail_bytes[1h], 3600*24) < 0

Notes:

• rate() and irate() should only be used with Counter types
• Time window selection affects result accuracy
• Use by clause for grouped aggregation