[
https://issues.apache.org/jira/browse/FLINK-17032?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17091491#comment-17091491
]
Canbin Zheng edited comment on FLINK-17032 at 4/24/20, 11:41 AM:
-----------------------------------------------------------------
{quote}Another property that I would consider desirable is that all resources
with the same type are grouped together.
{quote}
Hi [~chesnay]! My original thought is:
a) all the K8s resources of a Flink cluster should have the same prefix
${cluster-id}.
b) try best to have the same postfix for the resources with the same type, such
as *svc*, *config*.
I agree that all resources with the same type are grouped together is a better
idea, so, how about the following convention:
* The Deployment: ${clusterId}
* The internal Service: ${clusterId}-svc-internal
* The external Service: ${clusterId}-svc-external
* The Flink Configuration ConfigMap: ${clusterId}-config-flink
* The Hadoop Configuration ConfigMap: ${clusterId}-config-hadoop
* The JobManager Pod: ${clusterId}\-${random string}-${random string}
* The TaskManager Pod:
${clusterId}\-taskmanager-${currentMaxAttemptId}-${currentMaxPodId}
The reason why the name of the job manager pods does not contain a resource
type is that: when we create a Deployment, it creates a {{replicaset}} and add
a random string{{(*66cf4d99b5)* }}to the deployment name for the {{replicaset}}
name. The {{replicaset}} creates the pod(s). Replica set also adds another
random string to each {{pod}}.
Given that the job manager pods do not contain a resource type in their names,
we don't specify a resource type for the task manager pods as well.
WDYT?
was (Author: felixzheng):
{quote}Another property that I would consider desirable is that all resources
with the same type are grouped together.
{quote}
Hi [~chesnay]! My original thought is:
a) all the K8s resources of a Flink cluster should have the same prefix
${cluster-id}.
b) try best to have the same postfix for the resources with the same type, such
as *svc*, *config*.
I agree that all resources with the same type are grouped together is a better
idea, so, how about the following convention:
* The Deployment: ${clusterId}
* The internal Service: ${clusterId}-svc-internal
* The external Service: ${clusterId}-svc-external
* The Flink Configuration ConfigMap: ${clusterId}-config-flink
* The Hadoop Configuration ConfigMap: ${clusterId}-config-hadoop
* The JobManager Pod: ${clusterId}-${random string}-${random string}
* The TaskManager Pod:
${clusterId}-taskmanager-${currentMaxAttemptId}-${currentMaxPodId}
The reason why the name of the job manager pods does not contain a resource
type is that: when we create a Deployment, it creates a {{replicaset}} and add
a random string{{(*66cf4d99b5)* }}to the deployment name for the {{replicaset}}
name. The {{replicaset}} creates the pod(s). Replica set also adds another
random string to each {{pod}}.
Given that the job manager pods do not contain a resource type in their names,
we don't specify a resource type for the task manager pods as well.
WDYT?
> Naming convention unification for all the Kubernetes Resources
> --------------------------------------------------------------
>
> Key: FLINK-17032
> URL: https://issues.apache.org/jira/browse/FLINK-17032
> Project: Flink
> Issue Type: Improvement
> Components: Deployment / Kubernetes
> Affects Versions: 1.10.0
> Reporter: Canbin Zheng
> Priority: Minor
> Fix For: 1.11.0
>
>
> Currently, the naming rules are different among the Kubernetes resources we
> have created, the rules are as follows:
> # The Deployment: ${clusterId}
> # The internal Service: ${clusterId}
> # The external Service: ${clusterId}-rest
> # The Flink Configuration ConfigMap: flink-config-${clusterId}
> # The Hadoop Configuration ConfigMap: hadoop-config-${clusterId}
> # The JobManager Pod: ${clusterId}\-${random string}-${random string}
> # The TaskManager Pod:
> ${clusterId}\-taskmanager-${currentMaxAttemptId}-${currentMaxPodId}
> In the future, we would add other Kubernetes resources, and it would be
> better to have a unified naming convention for all of them.
> This ticket proposes the following naming convention:
> * The Deployment: ${clusterId}
> * The internal Service: ${clusterId}-internal-svc
> * The external Service: ${clusterId}-external-svc
> * The Flink Configuration ConfigMap: ${clusterId}-flink-config
> * The Hadoop Configuration ConfigMap: ${clusterId}-hadoop-config
> * The JobManager Pod: ${clusterId}\-${random string}-${random string}
> * The TaskManager Pod:
> ${clusterId}\-taskmanager-${currentMaxAttemptId}-${currentMaxPodId}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
