https://cloudacademy.com/research/multi-cloud-kubernetes-devops-cloud-academy-data-report-nov-18/

November 2018 | Cloud Academy Data Report No. 1

In the inaugural Cloud Academy data report, we have some interesting data to share around the multi-cloud and Kubernetes that might validate your perceptions of shifts in the cloud computing industry. Before we get to the juicy stuff, let’s start by discussing our mission with these data reports. Cloud Academy is the digital skills management platform that makes customized technical training manageable and measurable at scale. We deliver powerful solutions that enable our customers to align teams with important milestones on their transformation roadmaps and beyond. We’re enthusiastically making things like Cloud RosterCloud Catalog, and this research publicly available because we’re passionate about cloud computing and in a position to:

  • democratize our aggregated and anonymized data – and any insights we can glean from it – to the broader cloud computing community
  • empower organizations to make more intelligent decisions about the job roles they define as critical to creating a culture of collaboration and innovation
  • help individuals understand what the market is expecting from a skills perspective and focus training efforts accordingly
  • inspire discussion, interpretations, and hypotheses around data (if you have ideas we haven’t considered, please share them directly  or find us on Twitter @cloudacademy).

About the Data in this Report

Cloud Academy collects and analyzes upwards of 3,000 job descriptions each day for several cloud job roles based in the United States (we do have plans to expand). The data is then de-duplicated and analyzed. The technical job roles for which we collect job postings are: Cloud ArchitectDevOps EngineerFull-Stack EngineerNetwork ArchitectSecurity EngineerQA Engineer, and Data Engineer. This data report in particular leverages our job posts data warehouse and will reference measurements like:

Skill Relevancy: this is the association of a given technical skill to a job role based on the existence of a technology’s mention in a job post.
Skill Proximity: this is the frequency with which two or more skills appear within a category of job posts during a particular timeframe.

Proof that the World is Officially Multi-Cloud

Just 12 months ago, the concept of “multi-cloud” still seemed like mostly marketing hype. We even published a white paper that tried to separate fact from fiction. At AWS Summits this year, we heard lots of whispering that companies were using Azure for a subset of their applications and workloads. At Microsoft Ignite in September, people were very eager to share their initiatives by platform and most included both Azure and AWS. The anecdotes, which indicated that the world is indeed multi-cloud, seem to be backed up by data.

Let’s start to unpack the data around multi-cloud by looking at which combinations of public cloud platforms DevOps practitioners are being asked to have sufficient familiarity with.

What cloud platforms are DevOps professionals being asked to understand?

DevOps is in and of itself a mindset. Some argue that there is no such thing as a DevOps Engineer. Despite this, the role exists. The DevOps Engineer job role is one of the most sought-after roles in the United States cloud job market. The measure of skill relevancy for DevOps is generally a good indication of what technical decisions organizations have been made and what technologies organizations are using.

So what cloud platforms are DevOps professionals being asked to understand proficiently? The answer is no longer just AWS, and the answer seems to be changing rapidly.

AWS, Azure, and Google Cloud Skill Proximity – Data from a Quarter Ago
AWS, Azure, and Google Cloud Skill Proximity – Current Data

In our most recent data set, of those DevOps job posts that mention AWS, 42% also mention Azure. This figure has consistently trended upward over the past couple quarters.  A mere quarter ago, that figure was 28%. Either all skills on AWS and Azure are perfectly interchangeable (some are in concept, but this simply is not the case) or more likely the multi-cloud is here to stay and it is comprised primarily of two public clouds: AWS and Azure.  Google Cloud Platform (GCP) is mentioned in only 7% of the job postings that mention AWS and in 15% of those that mention Azure.

The data indicates that organizations seeking Azure talent typically also look for people familiar with AWS. Of job posts mentioning Azure, 79% also mention AWS (up 9% from two quarters ago). This seems to support the observation that most companies went all-in on AWS initially and have been increasingly looking to Azure for certain workloads or to support certain teams.

That leaves GCP firmly in third place. Very few companies seem to be seeking DevOps professionals who are skilled with GCP and not AWS or Azure as well. For each of the past six weeks, over 90% of job posts that mention GCP also mention AWS. During the same time period, over 60% of job posts that mention GCP also mention Azure.

Companies Aren’t Necessarily Buying the Argument that GCP is the Best Fit for Data Engineering

We’ve often heard that companies are using GCP because of the best-fit-for-data-science argument. We inspected Data Engineering job posts over the same time period. GCP didn’t crack the top 20 technical skills list for the job role. Both AWS and Azure did. You can view the entire current stack ranking and filter by job role here.

We’ll keep our eyes on multi-cloud trends over the coming months and quarters. For now, saying “we’re multi-cloud” safely means “we use AWS and Azure.”

For now, saying that 'we are multi-cloud' safely means 'we use AWS and Azure.'CLICK TO TWEET

Kubernetes May be Driving Multi-Cloud

While architecting a single application across multiple public clouds is likely still not a common practical use case, flexibly shifting applications and workloads from one public cloud to another has become much easier thanks to the rise of containerization. So what role does the leading modern container orchestration technology, Kubernetes, play in a company’s willingness to shift across and between the leading public clouds?

First, let’s look at Kubernetes’ usage in conjunction with the leading three public cloud providers in the US. Examining the Cloud Architect data set, we can see how frequently AWS, Azure, or GCP are mentioned in posts that lead with K8s.

Usage of Kubernetes and AWS, Azure, or Google Cloud – Current Data

We wanted to understand the proximity of K8s to the use of two or more public clouds. To achieve this, we segmented all Cloud Architect job posts into three categories based on an indication of use of exactly one, two, or three public cloud providers. We then segmented those posts into two buckets: one where K8s was a required skill for the role and another where it was not. We found a correlation; as the number of public clouds mentioned in a given Cloud Architect job post increases, the likelihood that K8s is required also increases. Of those Cloud Architect job posts that reference one cloud, just 9% mention K8s. The figure increases to 15% and 33% for two clouds and three clouds, respectively.

We didn’t want to limit our analysis to Cloud Architects, so we ran the same analysis on our DevOps Engineer data set. The trend we found is very similar. Of DevOps Engineer job posts where just one cloud is mentioned, K8s is found 21% of the time. For those where two clouds are mentioned, K8s is mentioned 30% of the time. Finally, in DevOps job posts where AWS, Azure, and Google Cloud Platform were mentioned, Kubernetes is referenced 37% of the time.

We don’t want to confuse correlation with causation here. It is very likely that for some organizations, the use of multiple public cloud providers is essentially necessitating a containerization strategy. At the same time, it is also likely that K8s is enabling organizations to more easily leverage multiple cloud vendors.

Our bottom line? Now is a good time to flesh out your knowledge of and experience with KubernetesAWS, and Azure. In our next monthly data report, we look forward to taking you into the world of skills in Data Engineering and big data analytics.

Multi-Cloud & Kubernetes: Cloud Academy November 2018 Data Report的更多相关文章

  1. 朱晔和你聊Spring系列S1E11:小测Spring Cloud Kubernetes @ 阿里云K8S

    有关Spring Cloud Kubernates(以下简称SCK)详见https://github.com/spring-cloud/spring-cloud-kubernetes,在本文中我们主要 ...

  2. Spring Cloud Config整合Spring Cloud Kubernetes,在k8s上管理配置

    1 前言 欢迎访问南瓜慢说 www.pkslow.com获取更多精彩文章! Kubernetes有专门的ConfigMap和Secret来管理配置,但它也有一些局限性,所以还是希望通过Spring C ...

  3. Springboot整合Spring Cloud Kubernetes读取ConfigMap,支持自动刷新配置

    1 前言 欢迎访问南瓜慢说 www.pkslow.com获取更多精彩文章! Docker & Kubernetes相关文章:容器技术 之前介绍了Spring Cloud Config的用法,但 ...

  4. spring cloud kubernetes之serviceaccount permisson报错

    spring boot项目引用spring-cloud-starter-kubernetes <dependency> <groupId>org.springframework ...

  5. Spring Cloud (Spring Cloud Stream)解析

    执行脚本目录 /bin windows 在其单独的目录 快速上手 下载并且解压kafka压缩包 运行服务 以Windows为例,首先打开cmd: 1.  启动zookeeper: bin\window ...

  6. On cloud, be cloud native

    本来不想起一个英文名,但是想来想去都没能想出一个简洁地表述该意思的中文释义,所以就用了一个英文名称,望见谅. Cloud Native是一个刚刚由VMware所提出一年左右的名词.其表示在设计并实现一 ...

  7. springboot+cloud 学习(五)统一配置中心 spring cloud config + cloud bus + WebHooks +RibbitMQ

    前言 微服务要实现集中管理微服务配置.不同环境不同配置.运行期间也可动态调整.配置修改后可以自动更新的需求,Spring Cloud Config同时满足了以上要求.Spring Cloud Conf ...

  8. Cloud Native Weekly | Kubernetes 1.13发布

    云原生一周精选 1——Kubernetes 1.13发布 2——Kubernetes首次出现重大安全漏洞 3——Docker和微软公司推出云原生应用的部署规范 4——谷歌推出beta版本的Cloud ...

  9. Netflix OSS、Spring Cloud还是Kubernetes? 都要吧!

    Netflix OSS是由Netflix公司主持开发的一套代码框架和库,目的是解决上了规模之后的分布式系统可能出现的一些有趣问题.对于当今时代的Java开发者们来说,Netflix OSS简直就是在云 ...

随机推荐

  1. 转:Ogre的八叉树场景管理器OctreeSceneManager

    上面是我绘制的一张图. 关于八叉树场景管理器主要需要关注两个类,其一是松散八叉树的数据结构Ogre::Octree,其二是八叉树场景管理器Ogre::OctreeSceneManager. 下面摘录图 ...

  2. 子类化QTreeWidgetItem实现增加Item的属性

    因为有需求是点击QTreeWidgetItem需要获取该Item的节点的相关属性,Item需要保存关联的属性,那么就需要扩展QTreeWidgetItem,当然,C++中扩展修改一个类或组件的方式就是 ...

  3. android 用代码画虚线边框背景

                   1.虚线画效果,可以使用Android中的xml来做. 2.直接上代码: <RelativeLayout android:id="@+id/coupon_ ...

  4. (原)使用tensorboard显示loss

    转载请注明出处: http://www.cnblogs.com/darkknightzh/p/7416551.html 参考网址: http://blog.csdn.net/jerry81333/ar ...

  5. CASE WHEN 及 SELECT CASE WHEN的用法(写了一坨烂代码发现两条sql就行了, 哎)

    转自:http://blog.sina.com.cn/s/blog_4c538f6c01012mzt.html Case具有两种格式.简单Case函数和Case搜索函数. 简单Case函数 CASE  ...

  6. Tensorflow中的name_scope和variable_scope

    Tensorflow是一个编程模型,几乎成为了一种编程语言(里面有变量.有操作......). Tensorflow编程分为两个阶段:构图阶段+运行时. Tensorflow构图阶段其实就是在对图进行 ...

  7. OpenCV 学习笔记 01 安装OpenCV及相关依赖库

    本次学习是基于Window10进行的.语言为python3. 1 与opencv相关的库简介 1.1 numpy numpy 是 OpenCV 绑定 python 时所依赖的库,此意味着numpy在安 ...

  8. 微信图片分享遇到 checkArgs fail, thumbData is invalid

    该问题主要是微信图片限制32K以内的原因,可将bmpToByteArray方法进行进行改写. 原方法是: /** * 得到Bitmap的byte * @author netcorner * @para ...

  9. AndroidStudio 编译异常java.lang.OutOfMemoryError: GC overhead limit exceeded

    在build.gradle中的android{}添加如下脚本就可以顺利编译了 dexOptions { incremental true javaMaxHeapSize “4g” }

  10. Python 文件 readline() 方法

    描述 Python 文件 readline() 方法用于从文件读取整行,包括 "\n" 字符.如果指定了一个非负数的参数,则返回指定大小的字符数,包括 "\n" ...