This is an automated email from the ASF dual-hosted git repository.
benjobs pushed a commit to branch dev
in repository
https://gitbox.apache.org/repos/asf/incubator-streampark-website.git
The following commit(s) were added to refs/heads/dev by this push:
new ef0e4d4d [Improve] Localization flink-development-framework-streampark
(#280)
ef0e4d4d is described below
commit ef0e4d4d45d1c6451c7f8675ed7c9bb2e220ef3e
Author: VampireAchao <[email protected]>
AuthorDate: Sun Oct 29 14:58:40 2023 +0800
[Improve] Localization flink-development-framework-streampark (#280)
* [Improve] Localization flink-development-framework-streampark
---
blog/1-flink-framework-streampark.md | 190 +++++++++++++++++------------------
1 file changed, 95 insertions(+), 95 deletions(-)
diff --git a/blog/1-flink-framework-streampark.md
b/blog/1-flink-framework-streampark.md
index dcd97dd7..b232658b 100644
--- a/blog/1-flink-framework-streampark.md
+++ b/blog/1-flink-framework-streampark.md
@@ -1,189 +1,189 @@
---
slug: flink-development-framework-streampark
-title: Flink 开发利器 StreamPark
+title: Flink Development Toolkit StreamPark
tags: [StreamPark, DataStream, FlinkSQL]
---
<br/>
-# 1. 背景
+# 1. Background
-Hadoop
体系虽然在目前应用非常广泛，但架构繁琐、运维复杂度过高、版本升级困难，且由于部门原因，数据中台需求排期较长，我们急需探索敏捷性开发的数据平台模式。在目前云原生架构的普及和湖仓一体化的大背景下，我们已经确定了将
Doris 作为离线数据仓库，将 TiDB（目前已经应用于生产）作为实时数据平台，同时因为 Doris 具有 on MySQL 的 ODBC
能力，所以又可以对外部数据库资源进行整合，统一对外输出报表
+Although the Hadoop system is widely used today, its architecture is
complicated, it has a high maintenance complexity, version upgrades are
challenging, and due to departmental reasons, data center scheduling is
prolonged. We urgently need to explore agile data platform models. With the
current prevalence of cloud-native architecture and the backdrop of lake and
warehouse integration, we have decided to use Doris as an offline data
warehouse and TiDB (which is already in production) as [...]
-<center style={{"color": "gray"}}>(这里借用一下 Doris 官方的架构图)</center>
+<center style={{"color": "gray"}}>(Borrowing Doris's official architecture
diagram here)</center>
<br/><br/>
-# 2. 遇到的问题
+# 2. Challenges Faced
-在数据引擎上，我们确定使用 Spark 和 Flink
+For the data engine, we settled on using Spark and Flink:
-* 使用 Spark on K8s client 客户端模式做离线数据处理
-* 使用 Flink on K8s Native-Application/Session 模式做实时任务流管理
+* Use Spark on K8s client mode for offline data processing.
+* Use Flink on K8s Native-Application/Session mode for real-time task stream
management.
-在这里，实际上有一些问题我们一直没有彻底解决：
+Here, there are some challenges we haven't fully resolved:
-用过 Native-Application 模式的朋友都知道，每提交一个任务，都需要打包新的镜像，提交到私有仓库，然后再调用 Flink Run 指令沟通
K8s，去拉取镜像运行 Pod。任务提交之后，还需要去 K8s 查看 log, 但是：
+Those who have used the Native-Application mode know that each time a task is
submitted, a new image must be packaged, pushed to a private repository, and
then the Flink Run command is used to communicate with K8s to pull the image
and run the Pod. After the task is submitted, you need to check the log on K8s,
but:
-1. 任务运行监控怎么处理？
-2. 使用 Cluster 模式还是 NodePort 暴露端口访问 Web UI？
-3. 提交任务能否简化打包镜像的流程?
-4. 如何减少开发压力？
+1. How is task runtime monitoring handled?
+2. Do you use Cluster mode or expose ports using NodePort to access Web UI?
+3. Can the task submission process be simplified to avoid image packaging?
+4. How can we reduce the pressure on developers?
<br/><br/>
-# 3. 解决问题的过程
+# 3. Solving the Challenges
-以上的这些其实都是需要解决的问题，如果单纯地使用命令行去提交每个任务，是不现实的，任务量大了，会变得不可维护。如何解决这些问题变成一个不得不面对的问题。
+All of the above are challenges that need addressing. If we rely solely on the
command line to submit each task, it becomes unrealistic. As the number of
tasks increases, it becomes unmanageable. Addressing these challenges became
inevitable.
<br/>
-## 简化镜像构建
+## Simplifying Image Building
-首先，针对 Flink 原生镜像需要二次 build 的问题：我们利用了 MinIO 作为外部存储，并使用 s3-fuse 通过 DaemonSet
的方式直接挂载在了每个宿主节点上，我们所需要提交的 jar 包都可以放到上面统一管理。这样的话，即使扩缩容 Flink 节点，也能实现 S3 挂载自动伸缩。
+Firstly, regarding the need for a secondary build of the native Flink image:
we utilized MinIO as external storage and mounted it directly on each host node
using s3-fuse via DaemonSet. The jar packages we need to submit can all be
managed there. In this way, even if we scale the Flink nodes up or down, S3
mounts can automatically scale.
-Flink 从 1.13 版本开始，就支持 Pod Template，我们可以在 Pod Template 中利用数据卷挂载的方式再将宿主机目录挂载到每个
pod 中，从而无需镜像打包而直接在 K8s 上运行 Flink 程序。如上图，我们将 S3 先通过 s3-fuse Pod 挂载在 Node 1、Node
2 的 `/mnt/data-s3fs` 目录下，然后再将 `/mnt/data-s3fs` 挂载到 Pod A 中。
+From Flink version 1.13 onwards, Pod Template support has been added. We can
use volume mounts in the Pod Template to mount host directories into each pod,
allowing Flink programs to run directly on K8s without packaging them into
images. As shown in the diagram above, we first mount S3 using the s3-fuse Pod
to the `/mnt/data-s3fs` directory on Node 1 and Node 2, and then mount
`/mnt/data-s3fs` into Pod A.
-但是，因为对象存储随机写入或追加文件需要重写整个对象，导致这种方式仅适合于频繁读。而这刚好满足我们现在的场景。
+However, because object storage requires the entire object to be rewritten for
random writes or file appends, this method is only suitable for frequent reads.
This perfectly fits our current scenario.
<br/>
-## 引入 StreamPark
+## Introducing StreamPark
-之前我们写 Flink SQL 基本上都是使用 Java 包装 SQL，打 jar 包，提交到 S3
平台上。通过命令行方式提交代码，但这种方式始终不友好，流程繁琐，开发和运维成本太大。我们希望能够进一步简化流程，将 Flink
TableEnvironment 抽象出来，有平台负责初始化、打包运行 Flink 任务，实现 Flink 应用程序的构建、测试和部署自动化。
+Previously, when we wrote Flink SQL, we generally used Java to wrap SQL,
packed it into a jar package, and submitted it to the S3 platform through the
command line. This approach has always been unfriendly; the process is
cumbersome, and the costs for development and operations are too high. We hoped
to further streamline the process by abstracting the Flink TableEnvironment,
letting the platform handle initialization, packaging, and running Flink tasks,
and automating the building, test [...]
-这是个开源兴起的时代，我们自然而然的将目光投向开源领域中：在一众开源项目中，经过对比各个项目综合评估发现 <span
style={{color:'red'}}> Zeppelin </span> 和 <span style={{color:'red'}}>
StreamPark </span> 这两个项目对 Flink 的支持较为完善，都宣称支持 <span style={{color:'red'}}>
Flink on K8s </span>，最终进入到我们的目标选择范围中，以下是两者在 K8s 相关支持的简单比较（目前如果有更新，麻烦批评指正）。
+This is an era of open-source uprising. Naturally, we turned our attention to
the open-source realm: among numerous open-source projects, after comparing
various projects, we found that both Zeppelin and StreamPark provide
substantial support for Flink and both claim to support Flink on K8s.
Eventually, both were shortlisted for our selection. Here's a brief comparison
of their support for K8s (if there have been updates since, please kindly
correct).
<table>
<thead>
<tr>
- <td>功能</td>
+ <td>Feature</td>
<td>Zeppelin</td>
<td>StreamPark</td>
</tr>
</thead>
<tbody>
<tr>
- <td>任务状态监控</td>
- <td><span style={{color:'red'}}> 稍低 </span>，不能作为任务状态监控工具</td>
- <td><span style={{color:'red'}}> 较高 </span></td>
+ <td>Task Status Monitoring</td>
+ <td>Somewhat limited, not suitable as a task status monitoring
tool.</td>
+ <td>Highly capable</td>
</tr>
<tr>
- <td>任务资源管理</td>
- <td><span style={{color:'red'}}> 无 </span></td>
- <td><span style={{color:'red'}}> 有 </span>，但目前版本还不是很健全</td>
+ <td>Task Resource Management</td>
+ <td>None</td>
+ <td>Exists, but the current version is not very robust.</td>
</tr>
<tr>
- <td>本地化部署</td>
- <td><span style={{color:'red'}}> 稍低 </span>，on K8s 模式只能将 Zeppelin
部署在 K8s 中，否则就需要打通 Pod 和外部网络，但是这在生产环境中很少这样做的</td>
- <td><span style={{color:'red'}}> 可以本地化部署 </span></td>
+ <td>Local Deployment</td>
+ <td>On the lower side. In on K8s mode, you can only deploy
Zeppelin in K8s. Otherwise, you need to connect the Pod and external network,
which is rarely done in production.</td>
+ <td>Can be deployed locally</td>
</tr>
<tr>
- <td>多语言支持</td>
- <td><span style={{color:'red'}}> 较高 </span>，支持 Python/Scala/Java
多语言</td>
- <td><span style={{color:'red'}}> 一般 </span>，目前 K8s 模式和 YARN 模式同时支持
FlinkSQL，并可以根据自身需求，使用 Java/Scala 开发 DataStream</td>
+ <td>Multi-language Support</td>
+ <td>High - Supports multiple languages such as
Python/Scala/Java.</td>
+ <td>Average - Currently, K8s mode and YARN mode support FlinkSQL,
and based on individual needs, you can use Java/Scala to develop
DataStream.</td>
</tr>
<tr>
- <td>Flink WebUI 代理</td>
- <td><span style={{color:'red'}}> 目前还支持的不是很完整 </span>，主开发大佬目前是考虑整合
Ingress</td>
- <td><span style={{color:'red'}}> 较好 </span>，目前支持
ClusterIp/NodePort/LoadBalance 模式</td>
+ <td>Flink WebUI Proxy</td>
+ <td>Currently not very comprehensive. The main developer is
considering integrating Ingress.</td>
+ <td>Good - Currently supports ClusterIp/NodePort/LoadBalance
modes.</td>
</tr>
<tr>
- <td>学习成本</td>
- <td><span style={{color:'red'}}> 成本较低 </span>，需要增加额外的参数学习，这个和原生的
FlinkSQL 在参数上有点区别</td>
- <td><span style={{color:'red'}}> 无成本 </span>，K8s 模式下 FlinkSQL
为原生支持的 SQL 格式；同时支持 Custome-Code（用户编写代码开发Datastream/FlinkSQL 任务）</td>
+ <td>Learning Curve</td>
+ <td>Low cost. Needs to learn additional parameters, which differ
somewhat from native FlinkSQL.</td>
+ <td>No cost. In K8s mode, FlinkSQL is supported in its native SQL
format; also supports Custom-Code (user writes code for developing
Datastream/FlinkSQL tasks).</td>
</tr>
<tr>
- <td>Flink 多版本支持</td>
- <td><span style={{color:'red'}}> 支持 </span></td>
- <td><span style={{color:'red'}}> 支持 </span></td>
+ <td>Support for Multiple Flink Versions</td>
+ <td>Supported</td>
+ <td>Supported</td>
</tr>
<tr>
- <td>Flink 原生镜像侵入</td>
- <td><span style={{color:'red'}}> 有侵入 </span>，需要在 Flink 镜像中提前部署 jar
包，会同 JobManager 启动在同一个 Pod 中，和 zeppelin-server 通信</td>
- <td><span style={{color:'red'}}> 无侵入 </span>，但是会产生较多镜像，需要定时清理</td>
+ <td>Intrusion into Native Flink Image</td>
+ <td>Invasive. You need to pre-deploy the jar package in the Flink
image, which will start in the same Pod as JobManager and communicate with the
zeppelin-server.</td>
+ <td>Non-invasive, but it will generate many images that need to be
cleaned up regularly.</td>
</tr>
<tr>
- <td>代码多版本管理</td>
- <td><span style={{color:'red'}}> 支持 </span></td>
- <td><span style={{color:'red'}}> 支持 </span></td>
+ <td>Multi-version Code Management</td>
+ <td>Supported</td>
+ <td>Supported</td>
</tr>
</tbody>
</table>
-<center style={{"color": "gray"}}>（PS: 此处仅从调研用户角度出发，我们对双方开发都保持极大的尊重）</center>
+<center style={{"color": "gray"}}>(PS: This comparison is based on our
perspective as evaluators. We hold the utmost respect for the developers of
both platforms.)</center>
<br/>
-调研过程中，我们与两者的主开发人员都进行了多次沟通。经过我们反复研究之后，还是决定将 StreamPark 作为我们目前的 Flink 开发工具来使用。
+During our research process, we communicated with the main developers of both
tools multiple times. After our repeated studies and assessments, we eventually
decided to adopt StreamPark as our primary Flink development tool for now.
<video src="http://assets.streamxhub.com/streamx-video.mp4";
controls="controls" width="100%" height="100%"></video>
-<center style={{"color": "gray"}}>(StreamPark 官网的闪屏)</center>
+<center style={{"color": "gray"}}>(StreamPark's official splash
screen)</center>
<br/>
-经过开发同学长时间开发测试，StreamPark 目前已经具备：
+After extended development and testing by our team, StreamPark currently
boasts:
-* 完善的<span style={{"color": "red"}}>SQL 校验功能</span>
-* 实现了<span style={{"color": "red"}}>自动 build/push 镜像</span>
-* 使用自定义类加载器，通过 Child-first 加载方式 <span style={{"color": "red"}}>解决了 YARN 和 K8s
两种运行模式</span>、<span style={{"color": "red"}}>支持了自由切换 Flink 多版本</span>
-* 与 Flink-Kubernetes 进行深度整合，提交任务后返回 WebUI，通过 remote rest api + remote
K8s，<span style={{"color": "red"}}>追踪任务执行状态</span>
-* 同时支持了 <span style={{"color": "red"}}>Flink 1.12、1.13、1.14 等版本</span>
+* Comprehensive <span style={{"color": "red"}}>SQL validation
capabilities</span>
+* It has achieved <span style={{"color": "red"}}>automatic build/push for
images</span>
+* Using a custom class loader and through the Child-first loading method, it
<span style={{"color": "red"}}>addresses both YARN and K8s operational
modes</span> and <span style={{"color": "red"}}>supports the seamless switch
between multiple Flink versions</span>
+* It deeply integrates with Flink-Kubernetes, returning a WebUI after task
submission, and via remote REST API + remote K8s, it can <span style={{"color":
"red"}}>track task execution status</span>
+* It supports versions like <span style={{"color": "red"}}>Flink 1.12, 1.13,
1.14, and more</span>
-以上基本解决了我们目前开发和运维中存在的大部分问题。
+This effectively addresses most of the challenges we currently face in
development and operations.
<video src="http://assets.streamxhub.com/streamx-1.2.0.mp4";
controls="controls" width="100%" height="100%"></video>
-<center style={{"color": "gray"}}>(StreamPark 对 Flink 多版本的支持演示视频)</center>
+<center style={{"color": "gray"}}>(Demo video showcasing StreamPark's support
for multiple Flink versions)</center>
<br/>
-在目前最新发布的 1.2.0 版本中，StreamPark 较为完善地支持了 K8s-Native-Application 和
K8s-Session-Application 模式。
+In its latest release, version 1.2.0, StreamPark provides robust support for
both K8s-Native-Application and K8s-Session-Application modes.
<video src="http://assets.streamxhub.com/streamx-k8s.mp4"; controls="controls"
width="100%" height="100%"></video>
-<center style={{"color": "gray"}}>(StreamPark K8s 部署演示视频)</center>
+<center style={{"color": "gray"}}>(StreamPark's K8s deployment demo
video)</center>
<br/>
-### K8s Native Application 模式
+### K8s Native Application Mode
-在 StreamPark 中，我们只需要配置相应的参数，并在 Maven POM 中填写相应的依赖，或者上传依赖 jar 包，点击
Apply，相应的依赖就会生成。这就意味着我们也可以将所有使用的 UDF 打成 jar 包，以及各种 connector.jar，直接在 SQL
中使用。如下图:
+Within StreamPark, all we need to do is configure the relevant parameters,
fill in the corresponding dependencies in the Maven POM, or upload the
dependency jar files. Once we click on 'Apply', the specified dependencies will
be generated. This implies that we can also compile all the UDFs we use into
jar files, as well as various connector.jar files, and use them directly in
SQL. As illustrated below:
-SQL 校验能力和 Zeppelin 基本一致:
+The SQL validation capability is roughly equivalent to that of Zeppelin:
-我们也可以指定资源，指定 Flink Run 中的动态参数 Dynamic Option，甚至参数可以整合 Pod Template
+We can also specify resources, designate dynamic parameters within Flink Run
as Dynamic Options, and even integrate these parameters with a Pod Template.
-程序保存后，点击运行时，也可以指定 savepoint。任务提交成功后，StreamPark 会根据 FlinkPod 网络 Exposed
Type（loadBalancer/NodePort/ClusterIp），返回相应的 WebURL，从而自然的实现 WebUI 跳转。但是，目前因为线上私有
K8s 集群出于安全性考虑，尚未打通 Pod 与客户端节点网络（目前也没有这个规划）。所以么，我们只使用 NodePort。如果后续任务数过多，有使用
ClusterIP 的需求的话，我们可能会将 StreamPark 部署在 K8s，或者同 Ingress 做进一步整合。
+After saving the program, when clicking to run, we can also specify a
savepoint. Once the task is successfully submitted, StreamPark will, based on
the FlinkPod's network Exposed Type (be it loadBalancer, NodePort, or
ClusterIp), return the corresponding WebURL, seamlessly enabling a WebUI
redirect. However, as of now, due to security considerations within our online
private K8s cluster, there hasn't been a connection established between the Pod
and client node network (and there's curre [...]
-注意：K8s master 如果使用 vip 做均衡代理的情况下，Flink 1.13 版本会返回 vip 的 ip 地址，在 1.14
版本中已经修复该问题。
+Note: If the K8s master uses a vip for load balancing, the Flink 1.13 version
will return the vip's IP address. This issue has been rectified in the 1.14
version.
-下面是 K8s Application 模式下具体提交流程
+Below is the specific submission process in the K8s Application mode:
-<center style={{"color": "gray"}}>（以上是依据个人理解绘制的任务提交流程图，如有错误，敬请谅解）</center>
+<center style={{"color": "gray"}}>(The above is a task submission flowchart,
drawn based on personal understanding. If there are inaccuracies, your
understanding is appreciated.)</center>
<br/>
-### K8s Native Session 模式
+### K8s Native Session Mode
-StreamPark 还较好地支持了 <span style={{"color": "red"}}> K8s Native-Sesson
模式</span>，这为我们后续做离线 FlinkSQL 开发或部分资源隔离做了较好的技术支持。
+StreamPark also offers robust support for the <span style={{"color": "red"}}>
K8s Native-Session mode</span>, which lays a solid technical foundation for our
subsequent offline FlinkSQL development or for segmenting certain resources.
-Native-Session 模式需要事先使用 Flink 命令创建一个运行在 K8s 中的 Flink 集群。如下：
+To use the Native-Session mode, one must first use the Flink command to create
a Flink cluster that operates within K8s. For instance:
```shell
./kubernetes-session.sh \
@@ -199,48 +199,48 @@ Native-Session 模式需要事先使用 Flink 命令创建一个运行在 K8s
-如上图，使用该 ClusterId 作为 StreamPark 的任务参数 Kubernetes ClusterId。保存提交任务后，任务会很快处于
Running 状态：
+As shown in the image above, we use that ClusterId as the Kubernetes ClusterId
task parameter for StreamPark. Once the task is saved and submitted, it quickly
transitions to a 'Running' state:
-我们顺着 application info 的 WebUI 点击跳转：
+Following the application info's WebUI link:
-可以看到，其实 StreamPark 是将 jar 包通过 REST API 上传到 Flink 集群上，并调度执行任务的。
+It becomes evident that StreamPark essentially uploads the jar package to the
Flink cluster through REST API and then schedules the task for execution.
<br/>
-### Custom Code 模式
+### Custom Code Mode
-另我们惊喜的是，StreamPark 还支持代码编写 DataStream/FlinkSQL 任务。对于特殊需求，我们可以自己写 Java/Scala
实现。可以根据 StreamPark 推荐的脚手架方式编写任务，也可以编写一个标准普通的 Flink 任务，通过这种方式我们可以将代码管理交由 git
实现，平台可以用来自动化编译打包与部署。当然，如果能用 SQL 实现的功能，我们会尽量避免自定义 DataStream，减少不必要的运维麻烦。
+To our delight, StreamPark also provides support for coding
DataStream/FlinkSQL tasks. For special requirements, we can author our
implementations in Java/Scala. You can compose tasks following the scaffold
method recommended by StreamPark or write a standard Flink task. By adopting
this approach, we can delegate code management to git, utilizing the platform
for automated compilation, packaging, and deployment. Naturally, if
functionality can be achieved via SQL, we would prefer not to [...]
<br/><br/>
-# 4. 意见和规划
+# 4. Feedback and Future Directions
-## 改进意见
+## Suggestions for Improvement
-当然 StreamPark 还有很多需要改进的地方，就目前测试来看：
+StreamPark, as with any new tool, does have areas ripe for enhancement based
on our current evaluations:
-* **资源管理还有待加强**：多文件系统jar包等资源管理功能尚未添加，任务版本功能有待加强。
-* **前端 button 功能还不够丰富**：比如任务添加后续可以增加复制等功能按钮。
-* **任务提交日志也需要可视化展示**：任务提交伴随着加载 class 文件，打 jar 包，build
镜像，提交镜像，提交任务等过程，每一个环节出错，都会导致任务的失败，但是失败日志往往不明确，或者因为某种原因导致异常未正常抛出，没有转换任务状态，用户会无从下手改进。
+* **Strengthening Resource Management**: Features like multi-file system jar
resources and robust task versioning are still awaiting additions.
+* **Enriching Frontend Features**: For instance, once a task is added,
functionalities like copying could be integrated.
+* **Visualization of Task Submission Logs**: The process of task submission
involves loading class files, jar packaging, building and submitting images,
and more. A failure at any of these stages could halt the task. Yet, error logs
often lack clarity, or due to some anomaly, the exceptions aren't thrown as
expected, leaving users puzzled about rectifications.
-众所周知，一个新事物的出现一开始总会不是那么完美。尽管有些许问题和需要改进的 point，但是瑕不掩瑜，我们仍然选择 StreamPark 作为我们的
Flink DevOps，我们也将会和主开发人员一道共同完善 StreamPark，也欢迎更多的人来使用，为 StreamPark 带来更多进步。
+It's a universal truth that innovations aren't perfect from the outset.
Although minor issues exist and there are areas for improvement with
StreamPark, its merits outweigh its limitations. As a result, we've chosen
StreamPark as our Flink DevOps platform. We're also committed to collaborating
with its main developers to refine StreamPark further. We wholeheartedly invite
others to use it and contribute towards its advancement.
<br/>
-## 未来规划
+## Future Prospects
-* 我们会继续跟进 Doris，并将业务数据 + 日志数据统一入 Doris，通过 Flink 实现湖仓一体；
-* 我们也会逐步将探索 StreamPark 同 DolphinScheduler 2.x 进行整合，完善DolphinScheduler 离线任务，逐步用
Flink 替换掉 Spark，实现真正的流批一体；
-* 基于我们自身在 S3 上的探索积累，fat-jar 包 build 完成之后不再构建镜像，直接利用 Pod Tempelet 挂载 PVC 到
Flink Pod 中的目录，进一步优化代码提交流程；
-* 将 StreamPark 持续应用到我们生产中，并汇同社区开发人员，共同努力，增强 StreamPark 在 Flink
流上的开发部署能力与运行监控能力，努力把 StreamPark 打造成一个功能完善的流数据 DevOps。
+* We'll keep our focus on Doris and plan to unify business data with log data
in Doris, leveraging Flink to realize lakehouse capabilities.
+* Our next step is to explore integrating StreamPark with DolphinScheduler
2.x. This would enhance DolphinScheduler's offline tasks, and gradually we aim
to replace Spark with Flink for a unified batch-streaming solution.
+* Drawing from our own experiments with S3, after building the fat-jar, we're
considering bypassing image building. Instead, we'll mount PVC directly to the
Flink Pod's directory using Pod Template, refining the code submission process
even further.
+* We plan to persistently implement StreamPark in our production environment.
Collaborating with community developers, we aim to boost StreamPark's Flink
stream development, deployment, and monitoring capabilities. Our collective
vision is to evolve StreamPark into a holistic stream data DevOps platform.
-附：
+Resources:
-StreamPark
GitHub：[https://github.com/apache/incubator-streampark](https://github.com/apache/incubator-streampark)
<br/>
-Doris GitHub：[https://github.com/apache/doris](https://github.com/apache/doris)
+StreamPark GitHub:
[https://github.com/apache/incubator-streampark](https://github.com/apache/incubator-streampark)
<br/>
+Doris GitHub:
[https://github.com/apache/doris](https://github.com/apache/doris)
