<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Amit Kumar Dube (अमित दुबे) on Medium]]></title>
        <description><![CDATA[Stories by Amit Kumar Dube (अमित दुबे) on Medium]]></description>
        <link>https://medium.com/@amitkumardube?source=rss-5fa603d6b074------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*et07-mVnvorDQCK4IA9Phw.jpeg</url>
            <title>Stories by Amit Kumar Dube (अमित दुबे) on Medium</title>
            <link>https://medium.com/@amitkumardube?source=rss-5fa603d6b074------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 08:09:47 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@amitkumardube/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[ctr vs nerdctl vs crictl]]></title>
            <link>https://amitkumardube.medium.com/ctr-vs-nerdctl-vs-crictl-8e3bece13bd6?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/8e3bece13bd6</guid>
            <category><![CDATA[k8s]]></category>
            <category><![CDATA[cli]]></category>
            <category><![CDATA[i̇mages]]></category>
            <category><![CDATA[cri]]></category>
            <category><![CDATA[containers]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Mon, 23 Dec 2024 16:49:46 GMT</pubDate>
            <atom:updated>2024-12-23T16:51:22.247Z</atom:updated>
            <content:encoded><![CDATA[<p>There are various command line tools to work with containers. As part of this post, we will talk about below 3 tools which works with containers, images and different container runtimes.</p><h3>ctr</h3><ul><li>The CLI is designed to work with containerd for debugging &amp; management.</li><li>Talks directly to containerd socket for full access to containers &amp; images.</li><li>Supports image pull, container management, streaming logs.</li><li>Included in containerd binary</li></ul><pre># Containerd<br>apt install containerd.io<br># CRI-O<br>apt install cri-o-tools</pre><h3>nerdctl</h3><ul><li>nerdctl provides easy-to-use docker compatible CLI that runs entirely in userspace.</li><li>Friendly UX for beginners starting with containers/Kubernetes.</li><li>Translates Docker CLI commands to containerd API calls.</li><li>Requires separate installation.</li><li>Works with containerd.</li></ul><pre># Containerd<br>apt install containerd.io<br><br>brew install nerdctl</pre><h3>crictl</h3><ul><li>It is used for CRI compatible container runtimes.</li><li>The tool is developed/maintained by k8s community.</li><li>It is more focused on pods more than containers/images.</li><li>crictl is included in the cri-tools package</li></ul><pre># Containerd<br>apt install containerd.io<br># CRI-O<br>apt install cri-o-tools</pre><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8e3bece13bd6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[GitHub Actions — Enable debug logging]]></title>
            <link>https://amitkumardube.medium.com/github-actions-enable-debug-logging-917aad0f6838?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/917aad0f6838</guid>
            <category><![CDATA[diagnosis]]></category>
            <category><![CDATA[debug]]></category>
            <category><![CDATA[logging]]></category>
            <category><![CDATA[github-actions]]></category>
            <category><![CDATA[github]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Wed, 03 Jan 2024 07:15:14 GMT</pubDate>
            <atom:updated>2024-01-03T07:18:22.322Z</atom:updated>
            <content:encoded><![CDATA[<h3>GitHub Actions — Enable debug logging</h3><p>When the GitHub workflow doesn’t provide enough information about why a workflow, job or step is not working as expected, then we enabled debug logging. Debug logging helps us to analyse an issue when the default logging is not sufficient.</p><p>The debug logging in GitHub Actions workflow can be enabled at 2 levels.</p><ul><li><strong>Runner debug logging </strong>— Provides more information about runner and worker.</li><li><strong>Step debug logging</strong> — Provides more information about workflow steps.</li></ul><p>The debug logging is enabled by creating a variable or secret in the repository that contains the GitHub Actions workflow.</p><h3>Important details</h3><ul><li>Follow this <a href="https://docs.github.com/en/actions/learn-github-actions/variables#creating-configuration-variables-for-a-repository">documentation</a> to create a repository variable.</li><li>Follow this <a href="https://docs.github.com/en/actions/security-guides/using-secrets-in-github-actions#creating-secrets-for-a-repository">documentation</a> to create a repository secret.</li><li>Follow this <a href="https://docs.github.com/en/actions/monitoring-and-troubleshooting-workflows/using-workflow-run-logs#downloading-logs">documentation</a> to access the GitHub workflow logs.</li></ul><h3><strong>Enable runner debug logging</strong></h3><ul><li>Set the value of secret or variable <strong>ACTIONS_RUNNER_DEBUG</strong> to <strong>true </strong>in the repository where workflow resides. Documentation on how to create a secret or variable is provided earlier.</li><li>If both secret and variable are set, the value of secret takes precedence.</li><li>Download the logs as per the steps in the documentation provided earlier. In addition to regular logs, there will now be an additional folder named <strong>runner-diagnostic-logs</strong>. This folder contains the debug logs for the <strong>worker</strong> and <strong>runner</strong>.</li><li>The <strong>runner logs</strong> provide information about setting up the runners to execute jobs.</li><li>The <strong>worker logs</strong> provide information about execution of a job.</li></ul><h3>Enable step debug logging</h3><ul><li>While debugging at runner level can be helpful in some cases, the step debugging helps us to get useful information about the execution of steps in the workflow.</li><li>Set the value of secret or variable <strong>ACTIONS_STEP_DEBUG</strong> to <strong>true </strong>in the repository where workflow resides. Documentation on how to create a secret or variable is provided earlier.</li><li>If both secret and variable are set, the value of secret takes precedence.</li></ul><figure><img alt="Step logs without debug logging enabled" src="https://cdn-images-1.medium.com/max/1024/1*JSXIEkqdu2Ve_p4POby1kg.png" /><figcaption><strong>Step logs without debug logging enabled</strong></figcaption></figure><figure><img alt="Step logs with debug logging enabled" src="https://cdn-images-1.medium.com/max/1024/1*j9VHlMO_Qmbk1utCuzx1Yg.png" /><figcaption><strong>Step logs with debug logging enabled</strong></figcaption></figure><p>Hope you find this article useful.</p><p>Happy Reading.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=917aad0f6838" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[GCP — small things — big savings]]></title>
            <link>https://amitkumardube.medium.com/gcp-small-things-big-savings-df95deaa573?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/df95deaa573</guid>
            <category><![CDATA[gcp]]></category>
            <category><![CDATA[money]]></category>
            <category><![CDATA[saving]]></category>
            <category><![CDATA[cloud]]></category>
            <category><![CDATA[cost]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Fri, 23 Jun 2023 05:48:33 GMT</pubDate>
            <atom:updated>2023-06-23T06:53:12.604Z</atom:updated>
            <content:encoded><![CDATA[<h3>GCP — small things — big savings</h3><p>Cost saving is a constant goal for any organization. While things are less panicking when setup is small but gradually as the setup grows, it becomes a real challege to control cost. There are few small things we could do to save cost while working with GCP.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*mEs27pOtMhn_Hk7l" /><figcaption><strong>Your saving is your earning</strong></figcaption></figure><blockquote>Your saving is your earning.</blockquote><h3>Delete Unused Reserved external IP addresses</h3><p><strong>$7.30 per month</strong> is charged by google for unused external reserved IP address.</p><p>So how to find unused IP addresses and delete them?</p><ul><li>You can go to IP address view in GCP.</li><li>The table has a column called “In use by”. If the value in this column is “None”, then this IP address is not used and can be deleted.</li><li>Please make sure that none of your external IP addresses are left unused.</li></ul><h3>Don’t reserve minimum no of instances when using cloud run if not needed</h3><ul><li>Cloud run supports the setup of minimum no of instance that will be running all the times. This is a feature to reduce the cold start.</li><li>While this is good features to avoid latency during initial calls, this could result in a lot of extra cost.</li><li>If your application is not a critical client facing application or if cold start of a few seconds is not a problem for you, then set minimum no of running instances to 0. <strong>This will save cost.</strong></li></ul><h3>Use 2nd generation cloud functions</h3><ul><li>2nd generation cloud function uses cloud run under the hood.</li><li>It supports concurrency which means that 1 instance can handle multiple requests.</li><li>With concurrency enabled, we can reduce the min instances (instances always running to avoid latency).</li><li><strong>This will reduce cost.</strong></li><li><a href="https://amitkumardube.medium.com/cloud-functions-2nd-gen-features-aa19a65026c2?source=friends_link&amp;sk=a3a666535640da8ce21a11d4af6ca209">Here</a> we can read about features of 2nd generation cloud functions.</li></ul><h3>If using Bigquery — Try Bigquery commitments</h3><ul><li><a href="https://cloud.google.com/bigquery/docs/reservations-commitments">https://cloud.google.com/bigquery/docs/reservations-commitments</a></li><li><strong>This will reduce cost.</strong></li></ul><h3>Avoid assigning external IP address to VM instances rather use load balancer to expose traffic outside of VPC</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2bTzrC4uYKai4o8abOuJng.png" /></figure><h3>Use Cloud Storage with GCSFUSE for small shared filesystem storage requirements instead of Filestore</h3><ul><li>If there is a need to have a filesystem storage attached on multiple instances, a cost effective alternative is to use <strong>cloud storage and mount is using GCSFUSE to the VMs</strong>.</li><li>There is a performance trade-off with this approach.</li><li>Ideally filestore is meant for this purpose but even a minimal filestore configuration will cost around 250 euros/month.</li><li>Using <strong>cloud storage with GCSFUSE</strong>, <strong>the cost is negligible.</strong></li></ul><h3>Network Cost due to egress traffic</h3><ul><li>Networks that use internal IPs in same zone to communicate can save google cloud egress bandwidth costs.</li><li>Check below table for more details.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*APohiuzwqVkJGfoq3HIN5Q.png" /></figure><h3>Logs Storage cost</h3><ul><li>The first <strong>50 GB / month is free</strong> in cloud logging.</li><li>Effective April 1, 2023, storage costs apply to logs data retained longer than 30 days. For pricing details, see <a href="https://cloud.google.com/stackdriver/pricing#logging-pricing-summary">Cloud Logging pricing summary</a>. To review the billable storage for your log buckets, go to the <strong>Logs Storage</strong> page of the Google Cloud console.</li><li>Be wise about what to log. It’s not free.</li></ul><h3>Network flow logs and firewall logs</h3><ul><li>Enabling firewall logging can generate a lot of logs. Please be wise while enabling them.</li><li>VPC flow logs keeps track of src_ip, src_port, dest_ip, dest_port and protocol. Enabling VPC flow logs can generate a lot of logs.</li><li>This will take your quota of free cloud logging storage and result in cost for storage beyond 50 GB/ month.</li></ul><h3>Audit Logs</h3><ul><li>Admin logs — enabled. 400 days. <strong>No charge.</strong></li><li>Data Access logs — should be enabled except for big query. 30 days default. Costs .50$ per GB data ingestion.</li><li>System Events — enabled. 400 days. <strong>No charge.</strong></li></ul><p><strong>Other</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d69PreZ16RZc5P3QncAvsg.png" /></figure><h3>Google Recommendations to reduce resources (CPU &amp; Memory)</h3><ul><li>Google provides recommendations for under utilised resources.</li><li>Go to <strong>cloud overview -&gt; Recommendations.</strong></li><li>There are recommendations about the unused resources and how much we can save if they are deleted.</li><li>There are recommendations about reducing image sizes and how much we can saved if machines are resized.</li><li>Please make a habit to visit this section at least once every month and act on them if needed.</li><li><strong>This will considerably save cost.</strong></li></ul><h3>Setup Horizontal autoscaling</h3><ul><li>We all are always used to over provision resources to have a better customer experience.</li><li>With GCP there are many compute services ie: cloud run, MIG, GKE, App Engine where scaling is done automatically for us without paying more and compromising one application performance.</li><li>Use horizontal autoscaling where possible <strong>to get the best of both worlds.</strong></li></ul><h3>Setup alerting and monitoring for underutilisation threshold if autoscaling is not setup</h3><ul><li>We all are always worried of setting threshold to CPU and memory usage of around 70–80 so that we are notified on time in order to scale things up.</li><li>However it’s not a common practice to setup threshold for underutilised resources. Like scaling down the memory and CPU when usage is below 40% for example.</li><li><strong>Setting such alerts can result in less cost.</strong></li></ul><h3>Don’t create load balancer for each service</h3><ul><li>Some teams try to create load balancers for each service/application to make is publicly accessible using a single static IP.</li><li>In GCP, we can use feature rich global http/s load balancer. They are proxy load balancers and can route traffic to different services/end points based on path rules and host rules. <strong>This can considerably save cost.</strong></li><li>They Can integrate with cloud IAP to abstract application authentication and authorisation for all apps and have common method for all the applications.</li><li>They can integrate with cloud armor to provide edge protection and backend protection.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=df95deaa573" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Terraform 1.5 — Things to notice]]></title>
            <link>https://awstip.com/terraform-1-5-things-to-notice-51e368e967ed?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/51e368e967ed</guid>
            <category><![CDATA[terraform]]></category>
            <category><![CDATA[check]]></category>
            <category><![CDATA[import]]></category>
            <category><![CDATA[strcontains]]></category>
            <category><![CDATA[generate-config-out]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Fri, 16 Jun 2023 06:49:04 GMT</pubDate>
            <atom:updated>2023-06-27T14:01:24.638Z</atom:updated>
            <content:encoded><![CDATA[<h3>Terraform 1.5 — Things to notice</h3><p>Terraform 1.5 is released this week with some interesting new features. As I am exploring these new features, I thought to put them in this blog post.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xc_uzxqHlkhio2od" /><figcaption>New Features in Terraform 1.5.0</figcaption></figure><h3>Upgrade Terraform to latest 1.5.0</h3><p>If you are on Mac, use brew package manager to upgrade terraform to latest version <strong>1.5.0</strong> .</p><pre>Amits-MacBook-Pro:frontend amitdube$ brew upgrade terraform<br><br>Amits-MacBook-Pro:frontend amitdube$ terraform version<br>Terraform v1.5.0</pre><h3>New Features</h3><h4>Check Block</h4><p>A new check block is introduced. The check block can validate your infrastructure outside the usual resource lifecycle. Check blocks address a gap between post-apply and functional validation of infrastructure.</p><ul><li><strong>Check</strong> block can access any data source from any provider in your terraform code.</li><li><strong>Check</strong> block can also have it’s own scoped data source which is only accessible to the check block.</li><li><strong>Check</strong> block is executed at the end of <strong>plan</strong> and <strong>apply</strong> and doesn’t block the terraform code execution. If check block fails then terraform will display warning with message specified as error_message in the assert block.</li></ul><p><strong>Successful execution</strong></p><pre>check &quot;status_code&quot; {<br>  data &quot;http&quot; &quot;google&quot; {<br>    url = &quot;https://www.google.com&quot;<br>  }<br><br>  assert {<br>    condition = data.http.google.status_code == 200<br>    error_message = &quot;${data.http.google.url} returned an unhealthy status code&quot;<br>  }<br>}<br><br>Amits-MacBook-Pro:frontend amitdube$ terraform apply<br>data.http.google: Reading...<br>data.http.google: Read complete after 1s [id=https://www.google.com]<br>Apply complete! Resources: 0 added, 0 changed, 0 destroyed.</pre><p><strong>Failed Execution (only warning)</strong></p><pre>check &quot;status_code&quot; {<br>  data &quot;http&quot; &quot;google_not_exist&quot; {<br>    url = &quot;https://google.com/notexist&quot;<br>  }<br><br>  assert {<br>    condition = data.http.google_not_exist.status_code == 200<br>    error_message = &quot;${data.http.google_not_exist.url} returned an unhealthy status code&quot;<br>  }<br>}<br><br>Amits-MacBook-Pro:frontend amitdube$ terraform apply<br>data.http.google_not_exist: Reading...<br>data.http.google_not_exist: Read complete after 0s [id=https://google.com/notexist]<br>╷<br>│ Warning: Check block assertion failed<br>│ <br>│   on main.tf line 7, in check &quot;status_code&quot;:<br>│    7:     condition = data.http.google_not_exist.status_code == 200<br>│     ├────────────────<br>│     │ data.http.google_not_exist.status_code is 404<br>│ <br>│ https://google.com/notexist returned an unhealthy status code<br>╵<br><br>Apply complete! Resources: 0 added, 0 changed, 0 destroyed.</pre><h4>Import Block</h4><p>Like to <a href="https://developer.hashicorp.com/terraform/language/modules/develop/refactoring#moved-block-syntax"><strong>moved</strong></a> block, terraform has now introduced the <strong>import</strong> block. Import block can replace the importing of resources via CLI. The import block can now be added to the code and executed using CI/CD workflows.</p><ul><li>Define the resource that was created outside terraform in your terraform configuration file.</li><li>Add the import block with ID of created source as the provider end and ID of resource in terraform code.</li><li>Run Terraform apply . This will import all the other resource attribute to terraform code and going forward this resource can be managed by terraform.</li></ul><pre><br>resource &quot;google_service_account&quot; &quot;cloudrun_sa&quot; {<br>}<br><br>import {<br>  id = &quot;projects/&lt;project_id&gt;/serviceAccounts/&lt;email&gt;&quot;<br>  to = google_service_account.cloudrun_sa<br>}</pre><h4><strong>Generate configuration for imported resources</strong></h4><p>Here is a good news around further simplification of import process in terraform. A new flag <strong>-generate-config-out=PATH </strong>is now added to <strong>terraform plan</strong> command.</p><ul><li>With import block added and missing corresponding terraform, terraform can create the configuration by running terraform plan with above mentioned flag.</li><li>Once configuration is created, it can be reviewed and merged with your main configuration.</li><li>This feature requires manual interruption in order to review the generated configuration, review it, adjust it and merge it.</li></ul><pre>import {<br>  id = &quot;projects/&lt;project_id&gt;/serviceAccounts/&lt;email&gt;&quot;<br>  to = google_service_account.cloudrun_sa<br>}<br><br>Amits-MacBook-Pro:frontend amitdube$ terraform plan -generate-config-out=generated.tf<br><br>Amits-MacBook-Pro:TWRO-Infra-bot amitdube$ cat generated.tf <br># __generated__ by Terraform<br># Please review these resources and move them into your main configuration files.<br><br># __generated__ by Terraform from &quot;projects/twro-infra-bot-9e79/serviceAccounts/cloudrun-backend@twro-infra-bot-9e79.iam.gserviceaccount.com&quot;<br>resource &quot;google_service_account&quot; &quot;cloudrun_sa&quot; {<br>  account_id   = &quot;cloudrun-backend&quot;<br>  description  = null<br>  disabled     = false<br>  display_name = &quot;Service Account to be used with cloudrun Application&quot;<br>  project      = &quot;&lt;project_id&gt;&quot;<br>  timeouts {<br>    create = null<br>  }<br>}</pre><h4>New Function strcontains</h4><ul><li>Returns <strong>true</strong> or <strong>false</strong> based whether a string is contained in another string.</li></ul><pre>Amits-MacBook-Pro:TWRO-Infra-bot amitdube$ terraform console<br>&gt; strcontains(&quot;hello world&quot;, &quot;world&quot;)<br>true<br>&gt; strcontains(&quot;hello world&quot;, &quot;universe&quot;)<br>false</pre><h4>plantimestamp function</h4><ul><li>Returns timestamp with terraform plan .</li></ul><pre>check &quot;terraform_io_certificate&quot; {<br>  data &quot;tls_certificate&quot; &quot;terraform_io&quot; {<br>    url = &quot;https://www.terraform.io/&quot;<br>  }<br><br>  assert {<br>    condition = timecmp(plantimestamp(), data.tls_certificate.terraform_io.certificates[0].not_after) &lt; 0<br>    error_message = &quot;terraform.io certificate has expired&quot;<br>  }<br>}</pre><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=51e368e967ed" width="1" height="1" alt=""><hr><p><a href="https://awstip.com/terraform-1-5-things-to-notice-51e368e967ed">Terraform 1.5 — Things to notice</a> was originally published in <a href="https://awstip.com/">AWS Tip</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[K8s Security — Use distroless base images]]></title>
            <link>https://amitkumardube.medium.com/k8s-security-use-distroless-base-images-854aa76ef1d3?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/854aa76ef1d3</guid>
            <category><![CDATA[containers]]></category>
            <category><![CDATA[security]]></category>
            <category><![CDATA[k8s]]></category>
            <category><![CDATA[distroless]]></category>
            <category><![CDATA[package]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Sat, 10 Jun 2023 17:34:17 GMT</pubDate>
            <atom:updated>2023-06-10T18:01:30.967Z</atom:updated>
            <content:encoded><![CDATA[<h3>K8s Security — Use distroless base images</h3><p>Security is fundamental and should be in the mind of everyone in an organization.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*g_xTI0Lkof7L3jlV" /><figcaption>K8s Security</figcaption></figure><h3>Important Points</h3><ul><li>syft — we are using this tool to list the packages inside an image.</li><li>We are comparing 3 base images. Debian, Alpine &amp; Distroless.</li></ul><h3>Debian Images</h3><ul><li>They contain lots of packages including <strong>apt</strong> package manager as listed below. These packages can be used to install other packages or to make remote calls.</li><li>The extent of the risk of using Debian-based images is plain to see: The more packages there are, the larger the attack surface becomes.</li></ul><pre>Amits-MacBook-Pro:k8s amitdube$ syft debian<br> ✔ Parsed image            <br> ✔ Cataloged packages      [96 packages]<br><br>NAME                    VERSION                       TYPE <br>adduser                 3.118                         deb   <br>apt                     2.2.4                         deb   <br>base-files              11.1+deb11u7                  deb   <br>base-passwd             3.5.51                        deb   <br>bash                    5.1-2+deb11u1                 deb   <br>bsdutils                1:2.36.1-8+deb11u1            deb   <br>coreutils               8.32-4+b1                     deb   <br>dash                    0.5.11+git20200708+dd9ef66-5  deb   <br>debconf                 1.5.77                        deb   <br>debian-archive-keyring  2021.1.1+deb11u1              deb   <br>debianutils             4.11.2                        deb   <br>diffutils               1:3.7-5                       deb   <br>dpkg                    1.20.12                       deb   <br>e2fsprogs               1.46.2-2                      deb   <br>findutils               4.8.0-1                       deb   <br>gcc-10-base             10.2.1-6                      deb   <br>gcc-9-base              9.3.0-22                      deb   <br>gpgv                    2.2.27-2+deb11u2              deb   <br>grep                    3.6-1+deb11u1                 deb   <br>gzip                    1.10-4+deb11u1                deb   <br>hostname                3.23                          deb   <br>init-system-helpers     1.60                          deb   <br>libacl1                 2.2.53-10                     deb   <br>libapt-pkg6.0           2.2.4                         deb   <br>libattr1                1:2.4.48-6                    deb   <br>libaudit-common         1:3.0-2                       deb   <br>libaudit1               1:3.0-2                       deb   <br>libblkid1               2.36.1-8+deb11u1              deb   <br>libbz2-1.0              1.0.8-4                       deb   <br>libc-bin                2.31-13+deb11u6               deb   <br>libc6                   2.31-13+deb11u6               deb   <br>libcap-ng0              0.7.9-2.2+b1                  deb   <br>libcom-err2             1.46.2-2                      deb   <br>libcrypt1               1:4.4.18-4                    deb   <br>libdb5.3                5.3.28+dfsg1-0.8              deb   <br>libdebconfclient0       0.260                         deb   <br>libext2fs2              1.46.2-2                      deb   <br>libffi7                 3.3-6                         deb   <br>libgcc-s1               10.2.1-6                      deb   <br>libgcrypt20             1.8.7-6                       deb   <br>libgmp10                2:6.2.1+dfsg-1+deb11u1        deb   <br>libgnutls30             3.7.1-5+deb11u3               deb   <br>libgpg-error0           1.38-2                        deb   <br>libgssapi-krb5-2        1.18.3-6+deb11u3              deb   <br>libhogweed6             3.7.3-1                       deb   <br>libidn2-0               2.3.0-5                       deb   <br>libk5crypto3            1.18.3-6+deb11u3              deb   <br>libkeyutils1            1.6.1-2                       deb   <br>libkrb5-3               1.18.3-6+deb11u3              deb   <br>libkrb5support0         1.18.3-6+deb11u3              deb   <br>liblz4-1                1.9.3-2                       deb   <br>liblzma5                5.2.5-2.1~deb11u1             deb   <br>libmount1               2.36.1-8+deb11u1              deb   <br>libnettle8              3.7.3-1                       deb   <br>libnsl2                 1.3.0-2                       deb   <br>libp11-kit0             0.23.22-1                     deb   <br>libpam-modules          1.4.0-9+deb11u1               deb   <br>libpam-modules-bin      1.4.0-9+deb11u1               deb   <br>libpam-runtime          1.4.0-9+deb11u1               deb   <br>libpam0g                1.4.0-9+deb11u1               deb   <br>libpcre2-8-0            10.36-2+deb11u1               deb   <br>libpcre3                2:8.39-13                     deb   <br>libseccomp2             2.5.1-1+deb11u1               deb   <br>libselinux1             3.1-3                         deb   <br>libsemanage-common      3.1-1                         deb   <br>libsemanage1            3.1-1+b2                      deb   <br>libsepol1               3.1-1                         deb   <br>libsmartcols1           2.36.1-8+deb11u1              deb   <br>libss2                  1.46.2-2                      deb   <br>libssl1.1               1.1.1n-0+deb11u4              deb   <br>libstdc++6              10.2.1-6                      deb   <br>libsystemd0             247.3-7+deb11u2               deb   <br>libtasn1-6              4.16.0-2+deb11u1              deb   <br>libtinfo6               6.2+20201114-2+deb11u1        deb   <br>libtirpc-common         1.3.1-1+deb11u1               deb   <br>libtirpc3               1.3.1-1+deb11u1               deb   <br>libudev1                247.3-7+deb11u2               deb   <br>libunistring2           0.9.10-4                      deb   <br>libuuid1                2.36.1-8+deb11u1              deb   <br>libxxhash0              0.8.0-2                       deb   <br>libzstd1                1.4.8+dfsg-2.1                deb   <br>login                   1:4.8.1-1                     deb   <br>logsave                 1.46.2-2                      deb   <br>lsb-base                11.1.0                        deb   <br>mawk                    1.3.4.20200120-2              deb   <br>mount                   2.36.1-8+deb11u1              deb   <br>ncurses-base            6.2+20201114-2+deb11u1        deb   <br>ncurses-bin             6.2+20201114-2+deb11u1        deb   <br>passwd                  1:4.8.1-1                     deb   <br>perl-base               5.32.1-4+deb11u2              deb   <br>sed                     4.7-1                         deb   <br>sysvinit-utils          2.96-7+deb11u1                deb   <br>tar                     1.34+dfsg-1                   deb   <br>tzdata                  2021a-1+deb11u10              deb   <br>util-linux              2.36.1-8+deb11u1              deb   <br>zlib1g                  1:1.2.11.dfsg-2+deb11u2       deb   <br>    </pre><h3>Alpine Images</h3><ul><li>They are considered as light weight and used for fast build and deploy.</li><li>Alpine Linux is a security-oriented, lightweight Linux distribution</li><li>However they still contains package manager that can be used to install other packages. This can cause security issues.</li></ul><pre>Amits-MacBook-Pro:k8s amitdube$ syft alpine<br> ✔ Parsed image            <br> ✔ Cataloged packages      [16 packages]<br><br>NAME                    VERSION      TYPE   <br>alpine-baselayout       3.4.3-r1     apk     <br>alpine-baselayout-data  3.4.3-r1     apk     <br>alpine-keys             2.4-r1       apk     <br>apk-tools               2.14.0-r0    apk     <br>busybox                 1.36.0       binary  <br>busybox                 1.36.0-r9    apk     <br>busybox-binsh           1.36.0-r9    apk     <br>ca-certificates-bundle  20230506-r0  apk     <br>libc-utils              0.7.2-r5     apk     <br>libcrypto3              3.1.0-r4     apk     <br>libssl3                 3.1.0-r4     apk     <br>musl                    1.2.4-r0     apk     <br>musl-utils              1.2.4-r0     apk     <br>scanelf                 1.3.7-r1     apk     <br>ssl_client              1.36.0-r9    apk     <br>zlib                    1.2.13-r1    apk</pre><h3>Distroless Images</h3><ul><li>They contain the bare minimum packages needed to use this image as base image.</li><li>They offer better security as <strong>they don’t contain package manager, shell or other packages</strong> and this reduces risk surface.</li><li>They contain CA certificate and /etc/passwd file. They also contain <strong>nonroot</strong> user that can be used to run process etc.</li></ul><pre>Amits-MacBook-Pro:k8s amitdube$ syft gcr.io/distroless/static-debian11<br> ✔ Parsed image            <br> ✔ Cataloged packages      [3 packages]<br>NAME        VERSION           TYPE <br>base-files  11.1+deb11u7      deb   <br>netbase     6.3               deb   <br>tzdata      2021a-1+deb11u10  deb   </pre><p>This is all for now. I propose to use distroless images rather than debian or alpine as your base image while building your application.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=854aa76ef1d3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cloud Functions 2nd gen features]]></title>
            <link>https://amitkumardube.medium.com/cloud-functions-2nd-gen-features-aa19a65026c2?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/aa19a65026c2</guid>
            <category><![CDATA[features]]></category>
            <category><![CDATA[cloudrun]]></category>
            <category><![CDATA[2nd-generation]]></category>
            <category><![CDATA[cloud-functions]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Thu, 01 Jun 2023 10:56:05 GMT</pubDate>
            <atom:updated>2023-06-01T10:57:16.470Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*JXB7QzsKAo-yY4jz" /><figcaption>Next Generation cloud functions</figcaption></figure><blockquote>Connect me on <a href="http://linkedin.com/in/amit-dube-%E0%A4%85%E0%A4%AE%E0%A4%BF%E0%A4%A4-%E0%A4%A6%E0%A5%81%E0%A4%AC%E0%A5%87-40b6351a">LinkedIn</a></blockquote><p>The 2nd generation of cloud functions come with below features.</p><h3>Long Processing time</h3><ul><li>Default timeout value for http based cloud functions is increased to <strong>60 minutes</strong> from earlier <strong>9 minutes</strong>.</li><li>Default timeout value of event driven cloud functions is increased to <strong>10 minutes</strong> from earlier <strong>9 minutes</strong>.</li></ul><h3>More Processing Power</h3><ul><li>Cloud functions now supports upto <strong>4 vcpu</strong> from earlier<strong> 2 vcpu.</strong></li><li>Cloud functions now supports upto <strong>16 GB</strong> memory from earlier <strong>8 GB</strong> of memory.</li></ul><h3>Traffic Splitting</h3><ul><li>The 2nd generation of cloud function uses cloud run under the hood.</li><li>Each new deployment of a cloud function create a new revision of corresponding cloud run instance.</li><li>Once we have more than 1 revisions of cloud run instance, traffic splitting can be implemented between them.</li><li>The traffic splitting is available at cloud run instance level and not at cloud function level.</li><li>Use below gcloud commands to list various revisions for a cloud run service and then update traffic to split it between different revisions.</li></ul><pre>gcloud run revisions list ......<br><br>gcloud run services update-traffic ......</pre><h3>Concurrency &amp; minimum instances</h3><ul><li>The 2nd generation of cloud function supports concurrency while the 1st generation could handle only 1 request / instance (supporting no concurrency).</li><li>This means that with 2nd generation, 1 instance can handle multiple requests.</li><li>The concurrency is supported at underlying cloud run instance level and can be set in there.</li><li>Since cloud functions supports setting min instances, we can save money by reducing min instances and adding concurrency.</li><li>Below command will set concurrency to 8.</li></ul><pre>gcloud run services update &lt;service_name&gt; --concurrency 8 .....</pre><h3>Event Integrations</h3><ul><li>Cloud function is now integrated with 90+ services using event arc and can be triggered from events generated from these services.</li><li>For example, we can write cloud function to send email after a query execution is completed in big query.</li></ul><p>This is all for now. Hope you will find it useful to understand key features of cloud functions 2nd gen.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aa19a65026c2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Refactoring — what & when]]></title>
            <link>https://amitkumardube.medium.com/refactoring-what-when-4830efd3065f?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/4830efd3065f</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[refactoring]]></category>
            <category><![CDATA[dry]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Mon, 29 May 2023 07:29:36 GMT</pubDate>
            <atom:updated>2023-05-30T06:01:29.083Z</atom:updated>
            <content:encoded><![CDATA[<h3>Refactoring — what &amp; when</h3><p>Refactoring is an important aspect of software development to keep the code relevant over time and improve it’s longevity. Without refactoring, a software tend to rotten and loose it’s descriptively over time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*McgP91xDvfb0DuQP" /><figcaption>Refactoring</figcaption></figure><h3>What is refactoring?</h3><p><strong>Refactoring is about</strong></p><blockquote>“changing the internal structure of software in order to make it easier to understand and reduce the cost of future modifications without changing its observable behaviour”.</blockquote><h3>When to refactor?</h3><p>While the general answer would be that it should be in our radar all the times when writing code. Refactoring and feature building should go hand in hand.</p><blockquote>While it’s a good idea to refactor code while developing a feature, we should keep them isolated when writing code. In essence we should not write code that refactors and add new features all at once. Ideally we should refactor, run tests to see if everything works, and then starts writing code for a new feature.</blockquote><h3>How to know if my code needs refactoring?</h3><p>Generally I look for one of the below 5 symptoms in the code that is the right candidate for refactoring.</p><blockquote><strong>Cryptic Variables, function or module names</strong></blockquote><p>This is a great starting point. If you see variables, functions and modules whose names are not aligned with what the code is doing then they should be renamed to better align with the intention of the code.</p><blockquote><strong>Duplicate Code</strong></blockquote><p>While we all try to follow the DRY (Don’t Repeat Yourself) principle when writing code, we always end up writing code where a logic is duplicated at multiple places. This is another candidate for refactoring. We should try to remove the duplicate code and point everything to one single copy of the code.</p><blockquote><strong>Long Function</strong></blockquote><p>We should always try to keep our functions short (&lt; 50 lines). Longer functions tend to violate single responsibility principle. Such functions should be avoided as much as possible.</p><blockquote><strong>Long Argument List</strong></blockquote><p>A lots of arguments/parameters make code difficult to understand and maintain. Instead use objects with good names to pass as parameters to functions. This improves the maintainability of the code.</p><blockquote><strong>Global Data</strong></blockquote><p>The last but not the least, is excessive usage of global data. The global data that is being modified could create a lot of unexpected behaviours. They are difficult to change for future requirements as their change can impact different parts of the program. If global variables are being used, try to reduce their scope as part of refactoring.</p><p>Happy reading &amp; refactoring.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4830efd3065f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Terraform 1.4 — Things to notice]]></title>
            <link>https://amitkumardube.medium.com/terraform-1-4-things-to-notice-ea107b0cda92?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/ea107b0cda92</guid>
            <category><![CDATA[upgrade]]></category>
            <category><![CDATA[iac]]></category>
            <category><![CDATA[news]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[terraform]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Mon, 03 Apr 2023 06:51:16 GMT</pubDate>
            <atom:updated>2023-05-23T11:11:24.756Z</atom:updated>
            <content:encoded><![CDATA[<h3>Terraform 1.4 — Things to notice</h3><p>Terraform 1.4 is released last month and brought some nice improvements. While I am playing around with this new version, I thought to describe them here and group them into categories.</p><h3>Security</h3><ul><li><strong>Sensitive values are now hidden in console input</strong></li></ul><pre>// creating non sensitive variable<br>variable &quot;checking_sensitive&quot; {<br>  type = string<br>}<br><br>// running terraform plan without supplying the default value<br>// value is visible on the console as variable is not marked as sensitive.<br>Amits-MacBook-Pro:terraform amitdube$ terraform plan -target terraform_data.test_data<br>var.checking_sensitive<br>  Enter a value: sdfdfsdf<br><br>// Marking the variable as sensitive<br>variable &quot;checking_sensitive&quot; {<br>  type = string<br>  sensitive = true<br>}<br><br>// As I am typing, the value is not visible on the console as the variable is sensitive.<br>Amits-MacBook-Pro:terraform amitdube$ terraform plan -target terraform_data.test_data<br>var.checking_sensitive<br>  Enter a value: </pre><ul><li><strong>Support of customer-managed GCP KMS keys to GCS backend to encrypt/decrypt the state file.</strong></li></ul><p>This adds additional layer of security if terraform state is sensitive which is the case in almost all the cases.</p><blockquote>To use customer-managed encryption keys, you need to create a key and give your project’s GCS service agent permission to use it with the Cloud KMS CryptoKey Encrypter/Decrypter predefined role.</blockquote><p>More details <a href="https://developer.hashicorp.com/terraform/language/settings/backends/gcs#customer-managed-encryption-keys-cloud-kms">here</a>.</p><pre>terraform {<br>  backend &quot;gcs&quot; {<br>    bucket  = &quot;tf-state-prod&quot;<br>    prefix  = &quot;terraform/state&quot;<br>    kms_encryption_key {<br>      location = &quot;us-east1&quot; # optional, defaults to &#39;global&#39;<br>      key_ring = &quot;terraform-keyring&quot;<br>      key = &quot;terraform-key&quot; <br>    }<br>  }<br>}</pre><ul><li><strong>A different service account can be used to access state bucket then the one used by terraform to deploy infra as per configuration files. This helps in implementing the concept of “least privilege implementation”.</strong></li></ul><pre>GOOGLE_BACKEND_IMPERSONATE_SERVICE_ACCOUNT=&quot;&lt;name of service account that will be used to connect to terraform state&gt;&quot;<br>GOOGLE_IMPERSONATE_SERVICE_ACCOUNT=&quot;&lt;name of the service account that will be used to deploy infrastructure&gt;&quot;</pre><h3>Automation</h3><ul><li><strong>Create terraform workspace if not already exist by setting a flag.</strong></li></ul><pre>// Currently I have only one workspace<br>Amits-MacBook-Pro:terraform amitdube$ terraform workspace list<br>* default<br><br>// If I try to switch to a namespace that doesnt exist. I will get an error.<br>Amits-MacBook-Pro:terraform amitdube$ terraform workspace select test<br><br>Workspace &quot;test&quot; doesnt exist.<br><br>You can create this workspace with the &quot;new&quot; subcommand <br>or include the &quot;-or-create&quot; flag with the &quot;select&quot; subcommand.<br><br>// I can use the -or-create flag to create the workspace if doesn&#39;t exist.<br>Amits-MacBook-Pro:terraform amitdube$ terraform  workspace select -or-create test<br>Created and switched to workspace &quot;test&quot;!<br><br>You&#39;re now on a new, empty workspace. Workspaces isolate their state,<br>so if you run &quot;terraform plan&quot; Terraform will not see any existing state<br>for this configuration.</pre><h3><strong>Replacement / New Resource</strong></h3><ul><li>There is an in built replacement of <strong>null_resource</strong> available now. It’s called <strong>terraform_data</strong> resource.</li><li>It can be used to trigger provisioners based on change in variables.</li><li>Additionally it can also store/calculate temporary values in input and output attributes.</li></ul><pre>variable &quot;checking_sensitive&quot; {<br>  type = string<br>  sensitive = true<br>}<br><br>resource &quot;terraform_data&quot; &quot;test_data&quot; {<br>    triggers_replace = var.checking_sensitive<br>    input = var.checking_sensitive<br>    provisioner &quot;local-exec&quot; {<br>      command = &quot;echo ${self.input} &gt;&gt; private_ips.txt&quot;<br>    }<br>}</pre><p><strong>Documentation </strong>— <a href="https://developer.hashicorp.com/terraform/language/resources/terraform-data">https://developer.hashicorp.com/terraform/language/resources/terraform-data</a></p><h3>New Command</h3><p>A new command is introduced to export terraform functions signature in machine-readable format.</p><ul><li>This command returns <strong>list of all the terraform functions</strong> along with their <strong>description</strong>, <strong>return type</strong> &amp; the <strong>list of parameters</strong>.</li></ul><pre>terraform metadata functions -json </pre><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ea107b0cda92" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cycle time — an evolutionary metric]]></title>
            <link>https://amitkumardube.medium.com/cycle-time-an-evolutionary-metric-69b1e8e075e3?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/69b1e8e075e3</guid>
            <category><![CDATA[metrics]]></category>
            <category><![CDATA[continuous-delivery]]></category>
            <category><![CDATA[evolution]]></category>
            <category><![CDATA[architecture]]></category>
            <category><![CDATA[cycle-time]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Sun, 19 Feb 2023 19:02:47 GMT</pubDate>
            <atom:updated>2023-02-19T19:02:47.994Z</atom:updated>
            <content:encoded><![CDATA[<h3>Cycle time — an evolutionary metric</h3><p>As we are moving to micro services architecture and agile way of working, we need to find a way to measure our engineering efficiency in order to evaluate our approach as early as possible in the software development cycle. Metrics are one way of such evaluation and one of such metric is <strong>cycle time.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*gaF5r_hjqqNadA1b" /><figcaption>Cycle time</figcaption></figure><p>Before we talk about this metric further, lets try to understand what cycle time is.</p><h3>What is cycle time?</h3><p>Cycle time is the time elapsed between the initiation and completion of a unit of work.</p><blockquote><strong>The cycle time starts when a developer starts working on a new feature and expires when that feature is running in a production environment.</strong></blockquote><blockquote><strong>Cycle time is the continuous delivery measure of engineering efficiency. </strong>Reduction of cycle time is one of the key goals of continuous delivery.</blockquote><h3>Why cycle time?</h3><ul><li>Cycle time as a continuous delivery measure is used to evaluate one of the characteristics of evalutionary architecture called incremental change as mentioned by Nead Ford in his book Building Evolutionary Architectures. The lesser the cycle time, it’s quicker to make incremental change.</li><li>On the contrary, if cycle time lengthens, it slows down how fast the project can deliver new generations, which affects architecture evolvability.</li><li>The evolution speed (velocity) is propotional to cycle time. The developers can’t evolve their systems faster than the project’s cycle time.</li><li>In other words, the faster teams can release software, the faster they can evolve parts of their system.</li></ul><p>Cycle time is therefore a critical metric in evolutionary architecture projects.</p><p>Hope this article helps to understand cycle time and use it as an important metric for your next project.</p><p>Happy Reading.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=69b1e8e075e3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Platform engineering — lets do it differently in 2023]]></title>
            <link>https://amitkumardube.medium.com/platform-engineering-lets-do-it-differently-in-2023-d898719153d4?source=rss-5fa603d6b074------2</link>
            <guid isPermaLink="false">https://medium.com/p/d898719153d4</guid>
            <category><![CDATA[platform-engineering]]></category>
            <category><![CDATA[internal-product]]></category>
            <category><![CDATA[2023]]></category>
            <category><![CDATA[platform-engineer]]></category>
            <category><![CDATA[differently]]></category>
            <dc:creator><![CDATA[Amit Kumar Dube (अमित दुबे)]]></dc:creator>
            <pubDate>Sat, 31 Dec 2022 20:09:34 GMT</pubDate>
            <atom:updated>2022-12-31T20:20:17.085Z</atom:updated>
            <content:encoded><![CDATA[<h3>Platform engineering — lets do it differently in 2023</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*9sYTuuBbrqXTb9aY" /><figcaption>Platform Engineering</figcaption></figure><blockquote><strong>Behind every successful organisation stand 3 failed and forgotten.</strong></blockquote><p>In last couple of years, I have been lucky to be part of some successful platform engineering journeys. Lately I also had an opportunity to talk to colleagues on different projects and understand their challenges around <strong>platform engineering.</strong></p><p>As we are binding this year, I thought to compile my findings from both experiences and share them here. I hope that these findings will guide you to a successful platform engineering journey.</p><h3>TL/DR</h3><ul><li>Treat platform as product</li><li>Consider product teams as clients</li><li>Avoid cognitive load for sysops/devops</li><li>Promote self service culture</li><li>Invest and Improve</li></ul><h3>Long Version</h3><h4><strong>Treat platform as Product</strong></h4><p>For every successful organisation, <strong>internal platform</strong> should be as important as any other <strong>external client facing</strong> applications. It’s a place where product teams host their solutions. So better <strong>internal platform</strong> will ultimately be foundation for better solutions by product teams. That’s why its important that we</p><ul><li>Apply product management techniques to <strong>internal platform</strong>.</li><li>Follow the organisation’s <strong>OKR cycle</strong> for <strong>internal platform</strong> as well.</li></ul><h4>Treat internal developers as customers</h4><p>In <strong>platform engineering</strong>, developers as the core users of the <strong>internal platform</strong> developed by platform engineers. That’s why it’s necessary to</p><ul><li>Treat developer’s/product team’s satisfaction a sign of success.</li><li>Let product teams understand the value of using the platform.</li><li>Establish empathy with internal developers.</li><li>Ensure regular collaboration with internal developers.</li><li>Improve the platform regularily based on the feedback from product teams.</li></ul><h4>Avoid cognitive load for sysops/devops</h4><p>For organisations who failed to achieve a successful adoption of platform engineering, the migration to cloud was considered a parallel additional activity for their existing sysops/devops team. Most of these colleagues had never written code and had very little knowledge around IaC concept. This additional expectation resulted in increase of cognitive load for them as they were expected to learn a lots of new thing while still managing their current workload.</p><ul><li>As mentioned earlier, consider platform engineering as separate product and accordingly setup a new team to work on this product.</li><li>Hire platform engineers who understand the core idea around platform engineering, self-service setup and can write IaC.</li><li>Don’t just move the existing devops or sysops teams completely to do platform engineering. However bringing few colleagues from these teams to new platform engineering team is a good idea.</li></ul><h4>Promote self service culture</h4><p>The <strong>platform engineering</strong> enables self-service capabilities for software engineering organisations. However it’s important that all the product teams should embrace this <strong>self service</strong> culture. Which means that</p><ul><li>Product teams should utilise the platform developed by platform engineering team to deliver solutions in a <strong>self service manner</strong>.</li><li>If there are <strong>limitations in current platform</strong>, then product teams should request the new features in order to continue using the platform without regular support from platform team.</li><li>Ideally platform team and product teams should talk regularly to improve the <strong>self service capability</strong> of the platform.</li><li>Every support provided by <strong>platform team</strong> to <strong>product teams</strong> should be followed by a follow-up jira ticket <strong>(a bug or a feature)</strong>.</li></ul><blockquote>An ideal <strong>internal platform</strong> is a <strong>platform</strong> where there are <strong>no tasks</strong> being logged by product teams for platform team. This means that product teams are able to carry out all the tasks by themselves in a self-service manner.</blockquote><h4>Invest and Improve</h4><p>One of the mistakes that organisations do is to cut down the platform budget as soon as the platform is ready and usable. Please note that getting the platform ready is just the beginning. The platform requires a continuous investment and improvement till it matures. In order to achieve this we need to</p><ul><li>Engage with internal developers to understand their experience.</li><li>Gather usage metrics.</li><li>Improve the platform regularly to maintain the ease of use.</li><li>Make sure to compile and present the business benefits from the platform during showcase meetings to receive regular funding.</li></ul><p>I hope that above findings will guide you to a successful platform engineering journey. Happy Reading.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d898719153d4" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>