Metrics
Nodegroup Metrics
Displays various metrics for a specific Nodegroup over a designated period, including:
- Target and Current Size: Shows the Nodegroup's target and current specifications. Current specifications indicate the number of units in normal service. When adding a new unit or increasing capacity, the condition of Target Size > Current Size might temporarily occur, which will later adjust to Target Size = Current Size. If Target Size != Current Size persists, it could suggest that the system is in an abnormal state, prompting you to contact the administrator.
- Utilization: Reflects the utilization rate of Nodegroup resources, factoring in CPU and memory consumption. If this rate consistently exceeds 80%, capacity expansion may be necessary.
- QPS: Indicates the number of SQL statements processed by Nodegroup each second, encompassing Select, Update, Insert, Delete, and Copy types.
- SQL Latency: P99 represents the execution time of 99% of SQL statements on Nodegroup; P90 reflects the execution time of 90%. If the P99 or P90 latency remains abnormal for an extended period (such as several minutes), an analysis in conjunction with business and system conditions is required.
- SQL Type Delays: P99 and P90 delays for various SQL types (select, insert, update, delete, copy) are recorded separately. If these indicators show long-term abnormalities (for example, several minutes), they should be analyzed alongside business and system conditions.
- Network Throughput: Shows the Nodegroup's network throughput, including the total bytes received and sent.
- Connections: Displays the total number of SQL connections on Nodegroup, categorizing them into active and idle connections.
- Failed Query Count: Represents the number of SQL statements that failed to execute per second in Nodegroup. A sudden increase in this value necessitates an analysis of business and system conditions.
- Affected Rows: Indicates the number of rows impacted by Nodegroup insert (INSERT), update (UPDATE), or delete (DELETE) operations per second. If any exceptions occur or the results diverge from expectations, an analysis in conjunction with business and system conditions is advised.
Database Metrics
Shows the storage sizes for the Nodegroup and Database dimensions:
- Storage size for each Nodegroup: Displays the total storage size for all databases within each Nodegroup.
- Storage size for individual databases: Displays the storage size for each separate database.
The database storage size encompasses the total storage space used by all data, including table indexes and transaction logs. Data ingestion, modification, indexing, transaction processing, schema changes, replication, and snapshots can all influence the overall storage size.