Insufficient exec-mem-pool-bytes allocation causes write slowdowns during Parquet persistence. Memory pressure during the gen1-duration interval (default 10 minutes) indicates the memory pool is undersized for write volume.
InfluxDB rejects points when field data types don't match the existing schema. A 400 response with field errors signals type conflicts between incoming data and stored data.
Queries timeout when filtering or grouping by high-cardinality dimensions. The number of unique tag value combinations explodes query execution cost, especially without proper partitioning.
Rising heap memory usage in InfluxDB correlates with growing application state size. When heap usage increases significantly, check checkpoint metrics to confirm state size growth and consider state cleanup or retention policies.
Uncontrolled disk growth in InfluxDB indicates missing or ineffective retention policies. Without automated data expiration, historical data accumulates indefinitely, driving storage costs and query performance degradation.
