Back up and restore StarRocks data — create a backup repository (S3 via broker), take full snapshots of tables/databases, check backup status, and restore from a snapshot. Use when setting up StarRocks backups, snapshotting tables before risky changes, configuring a backup repository, or recovering data from a snapshot.
Set up and manage StarRocks cluster topology — Frontend (FE), Backend (BE), and Broker roles, fe.conf/be.conf configuration, adding/removing/decommissioning nodes, and checking cluster status. Use when deploying or scaling a StarRocks cluster, configuring FE/BE nodes, adding followers/observers, or checking cluster health.
Load data into StarRocks using Stream Load (HTTP/real-time), Broker Load (batch from S3/HDFS), Routine Load (Kafka continuous), and INSERT INTO. Use when ingesting data into StarRocks, building load pipelines, mapping/transforming columns on load, checking load status, or troubleshooting data load failures.
Build StarRocks ETL pipelines and async/scheduled tasks — daily batch loads, incremental load with deduplication, real-time aggregation pipelines, SCD Type 2, and SUBMIT TASK for background/scheduled INSERT/CTAS execution with monitoring. Use when designing StarRocks ETL workflows, running long INSERT...SELECT in the background, scheduling recurring transformations, or backfilling data.
Create and manage StarRocks materialized views — synchronous rollups (auto-maintained), asynchronous MVs (scheduled refresh, joins, partitioned), transparent query rewrite, and MV refresh/management. Use when pre-aggregating data in StarRocks, speeding up repeated aggregations or joins, setting up scheduled MV refresh, or enabling automatic query rewrite.
Monitor and maintain a StarRocks cluster — health/metrics (backends, processlist, killing queries), compaction management, tablet repair/rebalancing, routine maintenance schedules, statistics collection for CBO, and troubleshooting commands. Use when checking StarRocks cluster health, diagnosing compaction or tablet issues, killing runaway queries, collecting statistics, or planning routine maintenance.
Optimize StarRocks queries — reading EXPLAIN/EXPLAIN VERBOSE, verifying partition pruning, choosing JOIN strategy (broadcast/shuffle/colocate), predicate pushdown, parallelism (pipeline_dop), index usage, and diagnosing slow queries via query profiles. Use when a StarRocks query is slow, when analyzing execution plans, or when tuning JOINs, pruning, or parallelism.
Manage StarRocks resources — session-level controls (query timeout, memory limit, parallelism, disk spill), resource groups for workload isolation (ETL vs dashboards), and storage volumes for remote/tiered storage. Use when isolating workloads, capping query memory/CPU, controlling parallelism, configuring resource groups, or setting up S3-backed storage volumes.