guoqinglong created FLINK-37942:
-----------------------------------
Summary: The rest.path configuration invalid in docker
apache/flink:2.0-scala_2.12-java11
Key: FLINK-37942
URL: https://issues.apache.org/jira/browse/FLINK-37942
Project: Flink
Issue Type: Bug
Environment: Docker: apache/flink:2.0-scala_2.12-java11
Jobmanager1:
flink-conf.yaml: |
scheduler-mode: reactive
# CPU,内存,磁盘和网络配置
jobmanager.memory.process.size: 1024m
# taskmanager.memory.process.size: 2048m
blob.server.port: 6124
jobmanager.rpc.port: 6123
taskmanager.rpc.port: 6122
process.working-dir: /opt/disk/data
state.checkpoints.dir: file:///opt/disk/checkpoint/flink-checkpoints
state.savepoints.dir: file:///opt/disk/checkpoint/flink-savepoints
# 日志存储位置配置
kubernetes.flink.log.dir: /opt/disk/log/app
env.log.dir: /opt/disk/log/app
web.log.path: /opt/disk/log/web
# taskmanager： 配置JobManager的信息
jobmanager.rpc.address: flink-jobmanager-job1
taskmanager.numberOfTaskSlots: 1
parallelism.default: 1
execution.checkpointing.interval: 10s
state.backend.local-recovery: true
state.backend.type: rocksdb
state.backend.async: true
state.backend.incremental: true
# rocksdb配置
state.backend.rocksdb.log.dir: /opt/disk/log/rocksdb
state.backend.rocksdb.log.file-num: 4
state.backend.rocksdb.rescaling.use-delete-files-in-range: true
# Web配置
rest.path:/job1
Jobmanager2:
flink-conf.yaml: |
scheduler-mode: reactive
# CPU,内存,磁盘和网络配置
jobmanager.memory.process.size: 1024m
# taskmanager.memory.process.size: 2048m
blob.server.port: 6124
jobmanager.rpc.port: 6123
taskmanager.rpc.port: 6122
process.working-dir: /opt/disk/data
state.checkpoints.dir: file:///opt/disk/checkpoint/flink-checkpoints
state.savepoints.dir: file:///opt/disk/checkpoint/flink-savepoints
# 日志存储位置配置
kubernetes.flink.log.dir: /opt/disk/log/app
env.log.dir: /opt/disk/log/app
web.log.path: /opt/disk/log/web
# taskmanager： 配置JobManager的信息
jobmanager.rpc.address: flink-jobmanager-job2
taskmanager.numberOfTaskSlots: 1
parallelism.default: 1
execution.checkpointing.interval: 10s
state.backend.local-recovery: true
state.backend.type: rocksdb
state.backend.async: true
state.backend.incremental: true
# rocksdb配置
state.backend.rocksdb.log.dir: /opt/disk/log/rocksdb
state.backend.rocksdb.log.file-num: 4
state.backend.rocksdb.rescaling.use-delete-files-in-range: true
# Web配置
rest.path:/job2
Reporter: guoqinglong
In k8s, the flink application mode is deployed. There are currently multiple
job managers. I want to set the rest.path configuration item to distinguish
different job manager services through different URL suffixes. The current
rest.path configuration is not effective.
For example, access jobmanager1 through http://localhost1:8081/job1/#/overview,
and access jobmanager2 through http://localhost2:8081/job2/#/overview.
Is this feature not available in the flink docker version? Or am I using it
incorrectly?
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
