Incremental Checkpointing not work

i have a yaml in k8s like below:

> flinkConfig:
jobmanager.heap.size: "2g"
jobmanager.execution.failover-strategy: region
jobmanager.web.log.path: /var/log/jobmanager/current
taskmanager.log.path: /var/log/taskmanager/current
taskmanager.exit-on-fatal-akka-error: true
taskmanager.network.memory.min: 128m
taskmanager.network.memory.max: 512m
taskmanager.network.memory.fraction: 0.125
taskmanager.maxRegistrationDuration: Inf
taskmanager.initial-registration-pause: 500ms
taskmanager.max-registration-pause: 5s
taskmanager.refused-registration-pause: 5s
taskmanager.heap.size: "2g"
web.upload.dir: /opt/flink
akka.framesize: 20m
parallelism.default: 1
state.backend: rocksdb
state.backend.incremental: true
state.backend.fs.checkpointdir: hdfs://.../checkpoints-backend
state.checkpoints.dir: hdfs://.../checkpoints
state.savepoints.dir: hdfs://.../savepoints
restart-strategy: fixed-delay
restart-strategy.fixed-delay.delay: 10000s
restart-strategy.fixed-delay.attempts: 6
savepointPath: hdfs://*/savepoints

but i found that Incremental Checkpointing didn't work, what's wrong with it.thanks for your replay

No source files or tests are identified. Start by checking the operator's handling of the shown flinkConfig and the RocksDB, HDFS checkpoint, and incremental-checkpoint settings, then reproduce the reported behavior with this Kubernetes configuration. Done means identifying whether the configuration is unsupported, ignored, or misconfigured and documenting the required change.

Written by the indexing model from the issue text.