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

Feat/ddp learner env

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!33

Feat/replicas failed

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!32

feat: add service informer

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!31

1. 请求/replicas/failed接口时如果删除的是aggregator,则需要把aggregator和ddp learner都重建
2. 将service的owner
controller改为其对应的pod
3. 统一将每个module的container name改为nervex-container
4. 统一将每个module的port
name改为nervex-port

1. add service informer
2. query service informer to generate urls for server http get requests

feat: create aggregator when learner request more than 1 gpu

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!29

Feat/http api aggregator

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!30

docs: add architecture

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!28

fix: fix bug that container status missmatch with pod phase when has one more containers

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!26

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

fix: fix bug that defer is not executed before return

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!25

当job succeeded之后defer没有执行是因为defer语句被放置在return之后了,所以需要把defer语句提到前面去

Fix/replicas volume not created

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!24

fix: fix bug that the terminating coordinator request server to create replicas

See merge request platform/CloudNative4AI/cluster-lifecycle/nervex-operator!23