1. implement getJobInfo and getNodeInfo
2. normalize log message
3. add service handler and use
subdomain to manage service
4. add jobStatusHandlers to handle events of job update
5. refactor
job transformation logic
6. update client-go to v0.21.5

1. move packages to pkg dir

1. update NerveXJob to DIJob
2. update group version to diengine.sensetime.com
3. update project
name to di-orchestrator

when a pod has one more containers, the pod is always in Running phase as well as there exists a
Running container. So wheen main container exits, but the sidecar is still Running, the pod will
keep in Running phase. we fix it

add sleep time after updating job status since the updates may have not been applied to cluster,
which influences the job status during testing

1. create aggregator by operator instead of server
2. improve code reuse rate
3. update crd