<?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 Richard Sands on Medium]]></title>
        <description><![CDATA[Stories by Richard Sands on Medium]]></description>
        <link>https://medium.com/@richard.sands?source=rss-72aea1791bfe------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*BmHFwEJyUPiXONzo.</url>
            <title>Stories by Richard Sands on Medium</title>
            <link>https://medium.com/@richard.sands?source=rss-72aea1791bfe------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 16:18:02 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@richard.sands/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[How our kubernetes journey saved us $1 million dollars]]></title>
            <link>https://medium.com/hackernoon/how-our-kubernetes-journey-saved-us-1-million-dollars-cede771f1f2b?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/cede771f1f2b</guid>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[productivity]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Fri, 04 Jan 2019 04:16:15 GMT</pubDate>
            <atom:updated>2019-01-04T04:16:15.721Z</atom:updated>
            <content:encoded><![CDATA[<p>Early in 2018 our engineering team embarked on a journey to revamp our infrastructure and tooling. In essence our <em>developer experience</em> was far from were we needed it to be.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*te5fg01CzSyPPAC8eew8ug.jpeg" /></figure><p>Having a poor developer experience led to: engineers making workarounds to sync their code changes to EC2’s directly bypassing the pipelines. Unit tests barely used as the pipeline took a long time to give feedback. Debugging wasn’t easy. If a critical bug was found on production, good luck getting the fix out in anything under 1 hour on a good day.</p><p>We wanted to do better, our goal was to…</p><blockquote>Move with speed and have pipelines deliver software in sub 5 minutes</blockquote><p>…and that is what started our kubernetes journey.</p><p><strong>TLDR</strong>: Want to save $1 million dollars AND have a happy development team? Provide the correct tooling.</p><h3>So did we really save $1 million dollars?</h3><p>Apparent click bait title aside, we can save that amount of money if we look at our delivery pipeline alone and forgetting our reduced infrastructure costs. We had all of our delivery pipelines running on Bamboo (an old Atlassian product). It looked something like this:</p><ul><li>Code push</li><li>Manually go and hit Run Build in Bamboo 🐌</li><li>Bamboo built an EC2 AMI 🐌</li><li>Spun up a load balancer, EC2, RDS database, Memcache, etc ($ and 🐌)</li></ul><p>The process took around 45 minutes on average (if the developer was lucky not to be queued up behind others builds).</p><p>We moved to kubernetes/circleci and it looks like this:</p><ul><li>Code push</li><li>Automatic build triggered</li><li>Unit tests run (1 minute) ⚡</li><li>Build a docker image and push (2 minutes) ⚡</li><li>Deploy to kubernetes cluster (1 minute) ⚡</li></ul><p>The difference, no extra EC2’s, no extra load balancers and a 4 minute timeline.</p><h3>What about that $1 million dollars</h3><p>Our team is approximately 28 engineers, lets say 25 of them do 5 builds per day, 265 days a year. Our builds are running 40 minutes faster. Quick calculation:</p><blockquote>5 builds x 40 minutes saving = 200 minutes x 25 engineers = 5,000 minutes x 265 days a year = 1,325,000 minutes / 60 = 22,083 hours x avg $50 per hour staff cost = <strong>$1.1 million dollars</strong></blockquote><p>I hear you say, “but engineers are not sitting waiting on builds to happen, they multi task”.</p><p>They do but looking back; It took a full 45 minutes to find out if a build worked when running on AWS, we didn’t have docker back then. Now engineers can be confident their tests pass, code works on a standardised pipeline and share development URLs with other teams within 4 minutes of pushing code.</p><p>Migrating to Circle CI/Kubernetes has given us a modern pipeline that the team can be proud off. The team can fix critical production issues in minutes, not hours.</p><h3>Have a goal in mind, provide the tooling to reach that goal and enable your engineers to move with speed.</h3><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cede771f1f2b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/how-our-kubernetes-journey-saved-us-1-million-dollars-cede771f1f2b">How our kubernetes journey saved us $1 million dollars</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How our Kubes/CI/CD journey saved us $1 million dollars]]></title>
            <link>https://medium.com/carsguide-engineering/how-our-kubernetes-journey-saved-us-1-million-dollars-2697d9f695f0?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/2697d9f695f0</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[development]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[docker]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Fri, 04 Jan 2019 04:14:29 GMT</pubDate>
            <atom:updated>2019-01-09T04:58:44.004Z</atom:updated>
            <content:encoded><![CDATA[<p>Early in 2018 our engineering team embarked on a journey to revamp our infrastructure and tooling. In essence our <em>developer experience</em> was far from were we needed it to be.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*te5fg01CzSyPPAC8eew8ug.jpeg" /></figure><p>Having a poor developer experience led to: engineers making workarounds to sync their code changes to EC2’s directly bypassing the pipelines. Unit tests barely used as the pipeline took a long time to give feedback. Debugging wasn’t easy. If a critical bug was found on production, good luck getting the fix out in anything under 1 hour on a good day.</p><p>We wanted to do better, our goal was to…</p><blockquote>Move with speed and have pipelines deliver software in sub 5 minutes</blockquote><p>…and that is what started our kubernetes journey.</p><p><strong>TLDR</strong>: Want to save $1 million dollars AND have a happy development team? Provide the correct tooling.</p><h3>So did we really save $1 million dollars?</h3><p>Apparent click bait title aside, we can save that amount of money if we look at our delivery pipeline alone and forgetting our reduced infrastructure costs. We had all of our delivery pipelines running on Bamboo (an old Atlassian product). It looked something like this:</p><ul><li>Code push</li><li>Manually go and hit Run Build in Bamboo 🐌</li><li>Bamboo built an EC2 AMI 🐌</li><li>Spun up a load balancer, EC2, RDS database, Memcache, etc ($ and 🐌)</li></ul><p>The process took around 45 minutes on average (if the developer was lucky not to be queued up behind others builds).</p><p>We moved to kubernetes/circleci and it looks like this:</p><ul><li>Code push</li><li>Automatic build triggered</li><li>Unit tests run (1 minute) ⚡</li><li>Build a docker image and push (2 minutes) ⚡</li><li>Deploy to kubernetes cluster (1 minute) ⚡</li></ul><p>The difference, no extra EC2’s, no extra load balancers and a 4 minute timeline.</p><h3>What about that $1 million dollars</h3><p>Our team is approximately 28 engineers, lets say 25 of them do 5 builds per day, 265 days a year. Our builds are running 40 minutes faster. Quick calculation:</p><blockquote>5 builds x 40 minutes saving = 200 minutes x 25 engineers = 5,000 minutes x 265 days a year = 1,325,000 minutes / 60 = 22,083 hours x avg $50 per hour staff cost = <strong>$1.1 million dollars</strong></blockquote><p>I hear you say, “but engineers are not sitting waiting on builds to happen, they multi task”.</p><p>They do but looking back; It took a full 45 minutes to find out if a build worked when running on AWS, we didn’t have docker back then. Now engineers can be confident their tests pass, code works on a standardised pipeline and share development URLs with other teams within 4 minutes of pushing code.</p><p>Migrating to Circle CI/Kubernetes has given us a modern pipeline that the team can be proud off. The team can fix critical production issues in minutes, not hours.</p><h3>Have a goal in mind, provide the tooling to reach that goal and enable your engineers to move with speed.</h3><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2697d9f695f0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/carsguide-engineering/how-our-kubernetes-journey-saved-us-1-million-dollars-2697d9f695f0">How our Kubes/CI/CD journey saved us $1 million dollars</a> was originally published in <a href="https://medium.com/carsguide-engineering">CAMS Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why vs What; Context is King]]></title>
            <link>https://medium.com/carsguide-engineering/why-vs-what-context-is-king-42a05ad3f408?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/42a05ad3f408</guid>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[purpose]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Wed, 19 Dec 2018 22:46:19 GMT</pubDate>
            <atom:updated>2018-12-19T22:57:08.642Z</atom:updated>
            <content:encoded><![CDATA[<p>One vital thing to make a productive product and engineering team is context and that context is king. Setting context has been a huge focus for us and a recent catch up with one of our engineers inspired me to write this article. I talked about <a href="https://engineering.carsguide.com.au/why-vs-why-now-96a97249a23c">why vs why now</a> when making engineering decisions but purpose is needed before we even get to the why.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*j7FcGUd-XKID1Jly7CE7LA.png" /></figure><p>Two questions, very different outcomes:</p><ul><li>Why are we doing feature A?</li><li>What are we trying to solve by doing this?</li></ul><p><em>The thought of these two questions has inspired me to try and set context in a slightly different way going forward that should help our product and engineering teams have a higher sense of purpose.</em></p><h3>Why are we doing feature A?</h3><p>Engineers were asked to do feature A and one of them, rightly so, asked this very question “why are we doing feature A”. Typical answers that engineers may face:</p><ul><li>“I said so”</li><li>“A senior manager/leader/executive said do it”</li><li>“We have a metric not were it needs to be, it needs to move”</li><li>“We want to improve conversions”</li></ul><p>It is easy to see that it gives no sense of ownership, no way for engineers to “buy in” to the problem to solve. Those types of responses are making your team become order takers ultimately leading to becoming unhappy.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/847/1*hkQSP4ZH0dm6insxT5Ol3w.png" /></figure><h3>Engineers are very talented people.</h3><p>They want to show their skills, they want to make a difference to the company. Many will have amazing ideas but are not sure how to convert them to meaningful business outcomes or how to convey their idea which leads to them saying nothing ever again.</p><p>To tap into this skill is not easy and as leaders it is our job to expose it for the good of the company and the engineers happiness.</p><blockquote>Change the question….</blockquote><h3><strong>What are we trying to solve by doing this?</strong></h3><p>Same as before but one engineer asks <strong>what </strong>are we trying to solve by doing this. Those answers before won’t fit. If anything they avoid, not answer, the question. Answers would have to flip to:</p><ul><li>“We want to improve our server response time by 200ms”</li><li>“Leads submitted improve by 20%”</li><li>“Convert 40% of our traffic to revenue generating traffic”</li><li>“Reduce bounce rate by 10%”</li><li>“Increase revenue by 10% by providing a messaging service”</li></ul><p>It is very clear what we are trying to solve.</p><h3>Accomplishment</h3><p>Take the answer “We want to improve conversions” from the why question. An engineer may understand conversions and completely buy into improving the conversion for you. But they will lack the feeling of accomplishment if they are not aware of <strong>what</strong> success looks like. Did their effort make a difference? Was the late night worth it?</p><p>By setting the context of <strong>what</strong> the team is trying to achieve we can;</p><ul><li>Flip ownership to the engineers</li><li>Trigger the hidden skill of engineers allowing them to suggest a better way</li><li>They can give 110% effort knowing what success will be</li><li>Sense of purpose when success is reached</li><li>And more importantly…a reason to come to work every day and be excited about making a difference</li></ul><h3>Conclusion</h3><p>Explain the what and follow on with the why. The why may be “a survey showed 40% of our users would like to be able to do X”. Align your team to the mission and they will do a lot for you. Set the right context, context is king.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=42a05ad3f408" width="1" height="1" alt=""><hr><p><a href="https://medium.com/carsguide-engineering/why-vs-what-context-is-king-42a05ad3f408">Why vs What; Context is King</a> was originally published in <a href="https://medium.com/carsguide-engineering">CAMS Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AWS re:Invent — Midnight madness launches Robomaker]]></title>
            <link>https://medium.com/hackernoon/aws-re-invent-midnight-madness-launches-robomaker-8a1fda4f85e1?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/8a1fda4f85e1</guid>
            <category><![CDATA[infrastructure]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[product]]></category>
            <category><![CDATA[servers]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Thu, 29 Nov 2018 03:40:25 GMT</pubDate>
            <atom:updated>2018-11-29T03:40:25.875Z</atom:updated>
            <content:encoded><![CDATA[<h3>AWS re:Invent — Midnight madness launches Robomaker</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IsmsvrFSAyrAE2oj3rSi0Q.jpeg" /></figure><p>Day 0 at AWS re:Invent and Amazon have kicked off the party/conference with an impressive midnight party. Not your normal party; a lot of grown adults eating chicken wings, a video commemorating said chickens, guinness world records for glow sticks and air drumming, popcorn, unlimited booze flowing way too easily (normal), start up launch and an AWS product launch.</p><h3>Start up launch; Hiber</h3><p>CTO/Co-founder, Maarten Engelen, from start up <a href="https://hiber.global/">Hiber</a> was at the event to launch their new product, a new way to communicate IoT devices globally using mini satellites. Ultra smart idea, low cost and you can order your own <a href="https://hiberband.myshopify.com/products/the-hiberkit%C2%AE-for-developers?variant=5417202450467\">IoT kit</a> to play around with. It will be interesting to watch how this opens up new ideas globally.</p><h3>New AWS product in town; Robomaker</h3><p>Robots are the future, there is no getting away from it. To start making robots is hard, it seems like you have to be nothing but a genius. Amazon have started their journey to solve the issues with <a href="https://aws.amazon.com/robomaker/">Robomaker</a>. You can simulate how your robot will work in AWS built environments setup by Amazon e.g. a living room or a retail store. Change parameters and re run thousands of simulations in parallel, something that would take days (and cost a lot) in the real world.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F7HGScNGah3E%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D7HGScNGah3E&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7HGScNGah3E%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4abd1093e2360a4e24eb2812cd74cbcd/href">https://medium.com/media/4abd1093e2360a4e24eb2812cd74cbcd/href</a></iframe><p>It is exciting and no doubt will open the way for many more startups to form quickly, easily and cheaply. BUT as someone mentioned beside me, nobody seemed that thrilled about it at midnight madness. Yes it is exciting for some but the crowd at the event didn’t seem to relate with the product. AWS re:Invent in a few years may have a slightly different attendance with robots flying around.</p><h3>Guinness World Records</h3><p>To add to the fun a few world records were attempted;</p><ul><li>Largest Air Drum Ensemble</li><li>Most People Lighting Glow Sticks Simultaneously</li><li>Largest chicken wing eating competition</li></ul><h3>Not forgetting the chickens…</h3><p>This one was a bit bizarre but there was a whole video dedicated to chickens whilst a chicken wing competition was under way. Take away what you wish from this….</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*a5u9G8QnQ0JZp5ffGvsc1Q.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8a1fda4f85e1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/aws-re-invent-midnight-madness-launches-robomaker-8a1fda4f85e1">AWS re:Invent — Midnight madness launches Robomaker</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AWS re:Invent — Midnight madness launches Robomaker]]></title>
            <link>https://medium.com/carsguide-engineering/aws-re-invent-midnight-madness-launches-robomaker-404fd89ac2f0?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/404fd89ac2f0</guid>
            <category><![CDATA[amazon]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[robots]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Tue, 27 Nov 2018 20:16:08 GMT</pubDate>
            <atom:updated>2018-11-27T20:16:08.878Z</atom:updated>
            <content:encoded><![CDATA[<h3>AWS re:Invent — Midnight madness launches Robomaker</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IsmsvrFSAyrAE2oj3rSi0Q.jpeg" /></figure><p>Day 0 at AWS re:Invent and Amazon have kicked off the party/conference with an impressive midnight party. Not your normal party; a lot of grown adults eating chicken wings, a video commemorating said chickens, guinness world records for glow sticks and air drumming, popcorn, unlimited booze flowing way too easily (normal), start up launch and an AWS product launch.</p><h3>Start up launch; Hiber</h3><p>CTO/Co-founder, Maarten Engelen, from start up <a href="https://hiber.global/">Hiber</a> was at the event to launch their new product, a new way to communicate IoT devices globally using mini satellites. Ultra smart idea, low cost and you can order your own <a href="https://hiberband.myshopify.com/products/the-hiberkit%C2%AE-for-developers?variant=5417202450467\">IoT kit</a> to play around with. It will be interesting to watch how this opens up new ideas globally.</p><h3>New AWS product in town; Robomaker</h3><p>Robots are the future, there is no getting away from it. To start making robots is hard, it seems like you have to be nothing but a genius. Amazon have started their journey to solve the issues with <a href="https://aws.amazon.com/robomaker/">Robomaker</a>. You can simulate how your robot will work in AWS built environments setup by Amazon e.g. a living room or a retail store. Change parameters and re run thousands of simulations in parallel, something that would take days (and cost a lot) in the real world.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F7HGScNGah3E%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D7HGScNGah3E&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7HGScNGah3E%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4abd1093e2360a4e24eb2812cd74cbcd/href">https://medium.com/media/4abd1093e2360a4e24eb2812cd74cbcd/href</a></iframe><p>It is exciting and no doubt will open the way for many more startups to form quickly, easily and cheaply. BUT as someone mentioned beside me, nobody seemed that thrilled about it at midnight madness. Yes it is exciting for some but the crowd at the event didn’t seem to relate with the product. AWS re:Invent in a few years may have a slightly different attendance with robots flying around.</p><h3>Guinness World Records</h3><p>To add to the fun a few world records were attempted;</p><ul><li>Largest Air Drum Ensemble</li><li>Most People Lighting Glow Sticks Simultaneously</li><li>Largest chicken wing eating competition</li></ul><h3>Not forgetting the chickens…</h3><p>This one was a bit bizarre but there was a whole video dedicated to chickens whilst a chicken wing competition was under way. Take away what you wish from this….</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*a5u9G8QnQ0JZp5ffGvsc1Q.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=404fd89ac2f0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/carsguide-engineering/aws-re-invent-midnight-madness-launches-robomaker-404fd89ac2f0">AWS re:Invent — Midnight madness launches Robomaker</a> was originally published in <a href="https://medium.com/carsguide-engineering">CAMS Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why vs Why now]]></title>
            <link>https://medium.com/carsguide-engineering/why-vs-why-now-96a97249a23c?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/96a97249a23c</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[product]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Tue, 20 Nov 2018 01:36:06 GMT</pubDate>
            <atom:updated>2018-11-20T03:44:10.813Z</atom:updated>
            <content:encoded><![CDATA[<p>I recently read an article that said teams should ask “why now” instead of “why” when making large product and engineering decisions. It struck a chord.</p><blockquote>There is a cost associated with all things software engineering, in house engineering is not an infinite resource.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H5DRPlxeHd_5127T5qH4Uw.jpeg" /><figcaption>Speed is of the essence when deciding what to build</figcaption></figure><h3>Why</h3><p>When the question of why we need a “feature” is asked I have found it is <em>usually </em>backed by two types of reasoning;</p><ul><li>Feature parity to competitor</li><li>Belief customers need said feature</li></ul><p>Engineering teams are busier than ever, corporate want to manage costs and see gains more than ever and this gives us a problem; we cannot build everything now.</p><h3>Why now</h3><p>There is fierce competition when it comes to a business that has a core function based on building software. Startups and small companies can sweep in quickly and rapidly which will have a costly consequence. Some thought provokers that you can dig into when deciding what to build now and prioritise</p><ul><li>What other work would be impacted if we built now</li><li>Is there a simpler solution that will serve the <em>core</em> purpose of the feature</li><li>Are we solving a problem that does not yet exist</li><li>Return on investment if done now vs 6 months out</li><li>Can we A/B test a “non working” example to gather predicted usage</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x82AmWtW6DnZC5O1A5oHMg.jpeg" /></figure><h3>Nobody likes to say no but if the rewards are small when building now vs delaying, the best answer may be to say no.</h3><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=96a97249a23c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/carsguide-engineering/why-vs-why-now-96a97249a23c">Why vs Why now</a> was originally published in <a href="https://medium.com/carsguide-engineering">CAMS Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why vs. Why now]]></title>
            <link>https://medium.com/hackernoon/why-vs-why-now-1c4f951cc7da?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/1c4f951cc7da</guid>
            <category><![CDATA[why-vs-why-now]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Tue, 20 Nov 2018 01:14:03 GMT</pubDate>
            <atom:updated>2018-11-20T17:25:14.304Z</atom:updated>
            <content:encoded><![CDATA[<p>I recently read an article that said teams should ask “why now” instead of “why” when making large product and engineering decisions. It struck a chord.</p><blockquote>There is a cost associated with all things software engineering, in house engineering is not an infinite resource.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H5DRPlxeHd_5127T5qH4Uw.jpeg" /><figcaption>Speed is of the essence when deciding what to build</figcaption></figure><h3>Why</h3><p>When the question of why we need a “feature” is asked I have found it is <em>usually </em>backed by two types of reasoning;</p><ul><li>Feature parity to competitor</li><li>Belief customers need said feature</li></ul><p>Engineering teams are busier than ever, corporate want to manage costs and see gains more than ever and this gives us a problem; we cannot build everything now.</p><h3>Why now</h3><p>There is fierce competition when it comes to a business that has a core function based on building software. Startups and small companies can sweep in quickly and rapidly which will have a costly consequence. Some thought provokers that you can dig into when deciding what to build now and prioritise</p><ul><li>What other work would be impacted if we built now</li><li>Is there a simpler solution that will serve the <em>core</em> purpose of the feature</li><li>Are we solving a problem that does not yet exist</li><li>Return on investment if done now vs 6 months out</li><li>Can we A/B test a “non working” example to gather predicted usage</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x82AmWtW6DnZC5O1A5oHMg.jpeg" /></figure><h3>Nobody likes to say no but if the rewards are small when building now vs delaying, the best answer may be to say no.</h3><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1c4f951cc7da" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/why-vs-why-now-1c4f951cc7da">Why vs. Why now</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Launching a product on time]]></title>
            <link>https://medium.com/carsguide-engineering/launching-a-product-on-time-4fb26a42d95f?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/4fb26a42d95f</guid>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[team]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Mon, 12 Nov 2018 05:55:56 GMT</pubDate>
            <atom:updated>2018-11-12T05:55:56.098Z</atom:updated>
            <content:encoded><![CDATA[<p>The engineering team at carsguide/Autotrader recently <a href="https://engineering.carsguide.com.au/launching-autotrader-91f943a4b6e2">launched a new product</a> on time, on budget and with a larger feature set than the business was sold. It was a ground up build from a blank page with a brand new platform under pinning it. Did the team just do the impossible?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BVupbEqYzpUf1cf9LMrd9A.jpeg" /></figure><p>There are <strong>a lot</strong> of challenges when building software; scope creep, unknown requirements, working with external service providers, architectural challenges and managing the business. We faced every single one of these. Some of my takeaways from it;</p><h3>Simplistic estimation</h3><p>This may sound crazy (it was crazy when I think about it) but t-shirt sizes. In a nutshell we broke out the requirements into “themes” of work. That may be “user login”, “submit an enquiry” or “search for a car”. The tech leads along with engineers shouted out small, medium or large in a few seconds. Whatever was said within those seconds was what we used. Why it maybe worked; it is easy for the human mind to think in something it is familiar with and visualise the difference, not how many days or hours a feature may take.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*kz4aANxQmIRS1S8MzyPUpA.jpeg" /><figcaption>T-shirt sizing a project</figcaption></figure><p>We estimated with speed, we trusted the instinct of our engineers and their past experiences building a product. If someone really disagreed we would go a size larger.</p><h3>Why why why</h3><p>In past projects our engineers had work given with a lot of detail, everything was thought off before hand and the engineers asked questions during planning sessions. This project didn’t quite have that, leads were not sure on the exact details for every piece of work.</p><blockquote>Business logic would change as we adapted to new knowledge during the build.</blockquote><p>What we once thought was correct was incorrect. Architecture changed how we built things, learning as we progressed changed things. However what the senior tech team did know was the why;</p><ul><li>Why our users would use a feature</li><li>Why it mattered to the business</li><li>Why it made sense to invest engineering time into it</li></ul><p><strong>Takeaway</strong>; never stop explaining the why. Engineers are smart people, they may know a better way of executing if you tell them the why.</p><h3>Pressure</h3><p>This is something that everyone struggles with, myself included, and there is no magic fix. We all react to pressure in different ways and we all cope with pressures in different ways to strike a <a href="https://engineering.carsguide.com.au/work-life-balance-3acc93fc18f6">happy life/work balance</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/658/1*AGux0df-YQMRN__attBcwg.png" /><figcaption>Keep your team focused</figcaption></figure><p><strong>Reduce the pressure on your team.</strong> Small bite size pieces of work keep the team focused. Our teams could have created thousands of Jira tickets but that would make engineers overloaded with information that is not relevant. For example our platform team never had more than 20–30 tickets in Jira. The leads kept the team focused on the work at hand and removed blockers.</p><p>Some may think that creates a potential bottleneck (if someone is limiting the work flowing to the team), but I look at this as gate keeping. Put someone in this role that is happy to say no, question the why, keep focus on the product vision and<em> protect the team</em>.</p><p>The side effect of doing this is the gate keeper becomes over loaded. The gate keeper has to context switch a lot. Context switching constantly is a skill that not everyone has but if you want your team to succeed you need to remove the noise and keep the focus. (Context switching post coming up soon).</p><p><em>Personally I love pressure and the satisfaction from having the engineering team give 100% and deliver the project on time made it worthwhile.</em></p><h3>Architectural decisions and “style”</h3><p>Not to end on a boring note but I feel this is a key part of any new product build. I am not overly excited with someone making a 50 page document detailing all architectural decisions but it is important that a team work in harmony with regards to architecture and coding style. More important though is to adapt a system as your knowledge of the project grows.</p><p>Something isn’t fit for purpose anymore; re-architect it. Always keep an eye on technical debt, spend time going back over pieces already built. If you do you will reduce the complexity of a system and build for speed</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4fb26a42d95f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/carsguide-engineering/launching-a-product-on-time-4fb26a42d95f">Launching a product on time</a> was originally published in <a href="https://medium.com/carsguide-engineering">CAMS Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Going all in with Kubernetes (Part 2)]]></title>
            <link>https://medium.com/carsguide-engineering/going-all-in-with-kubernetes-part-2-9d6af24d5676?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/9d6af24d5676</guid>
            <category><![CDATA[deployment]]></category>
            <category><![CDATA[infrastructure]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[docker]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Tue, 21 Aug 2018 06:08:14 GMT</pubDate>
            <atom:updated>2018-08-21T06:53:35.938Z</atom:updated>
            <content:encoded><![CDATA[<p>If you haven’t already read <a href="https://engineering.carsguide.com.au/going-all-in-with-kubernetes-part-1-1823e51ce07f">part 1 of the series</a>, our engineering team have gone all in with Kubernetes and I can comfortably say we are well on the way to having a production ready infrastructure. There is certainly a lot of noise in the industry right now around Kubernetes however when chatting to others there is a real lack of adoption. Some key themes seem to be emerging from these conversations;</p><ul><li>Red tape in a business due to unknown risk factors</li><li>Lack of use cases proving it works and works well</li><li>Lack of official AWS support (until recently)</li><li><strong>Fear</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*7xvuFm7pVmv1fRpNulTyTg.jpeg" /><figcaption>Kubernetes in the early stages, fear takes over</figcaption></figure><p>Some have asked why did we chose Kubernetes, why not use something like AWS ECS. Mainly we got it through the business on;</p><ul><li>Ground up rebuild of our technology, set us up for the future, not today</li><li>Cost savings e.g. ECS requires load balancers per service</li><li>Cloud provider agnostic</li><li>Let’s be honest; it is a shiny new tech and engineers love new tech</li></ul><h3>Fear</h3><p>We had chatted to a solution architect who had suggested to a number of companies to run Kubernetes and they all rejected the suggestion to implement it, they were just not comfortable with the risks. This is originally what sparked my interest in Kubernetes, what was so risky that nobody would roll it out to production. Risk = fear.</p><blockquote>Fear can either control us or we can chose to overcome the fear and control it.</blockquote><p>If nobody takes risks, we do not have the opportunity to learn and grow.</p><h3>Mitigating Fear and Focus</h3><p>There is certainly still some fear in our engineering team. Some may feel Kubernetes is a beast “out of control” and to some extent I would agree, but that is only because Kubernetes has automated the things we had to do code before.</p><p>Failure of a container, it recovers. Failure of a deployment, it recovers. Scaling of containers, done. Scaling of nodes, done. Handling DNS, done. The list goes on, Kubernetes does it for us.</p><p>I myself had many sleepless days (and nights) upskilling. Embracing the technology, choosing not to let fear win, I am now confident in the tech.</p><p>However, that said we have to mitigate that fear and ensure the whole engineering team is embracing the technology. The main thing we have done so far is embedded operations into our platform team enabling them to work closely with engineers. Engineers can leverage the operations skill set and vice versa. Our staging cluster is an exact mirror of our future production cluster, we put full production workloads on this cluster to ensure it will work as expected. What we are seeing;</p><blockquote>Our teams can focus on meaningful work not the low hanging fruit that Kubernetes abstracts and does well.</blockquote><h3><strong>Start small</strong></h3><p>When we started the journey we put our new platform straight onto Kubernetes. The risk was extremely low, the platform is not used in production. Should it fail we can quickly change direction. Had we asked the business to move our main revenue driving website onto Kubernetes first I do not think we would be on the same journey. The business has seen over the past few months Kubernetes can be trusted and when we launch our new brand to market in late 2018 we can further the case to move current websites over to Kubernetes.</p><h3>Dollars, dollars</h3><p>One of the biggest benefits we have seen so far is the potential cost savings. The team have already proven that Kubernetes can cope with 8 times our current traffic with only 6 nodes (EC2s). We expect to be able to run our whole production and staging cluster at 1/3 of the cost of our current setup.</p><h3>People cost too</h3><p>Dollars is not just about our hosting costs, peoples time is valuable and that is also a cost factor. Our teams have full control of the deployment life cycle, they monitor the services they deploy themselves. Our teams output has been higher than ever before, our teams are no longer waiting on operations to deploy a service, they are in full control.</p><p>A member of the team recently said “<em>(the) </em>tech team are moving faster than ever this year” and I cannot agree more. I firmly believe our choice in Kubernetes and the new tooling we now have has enabled us to move with speed.</p><h3>Part 3</h3><p>Part 3 is coming up in a few weeks, stay tuned were I will discuss how we have set our clusters up and the tools we used to do it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9d6af24d5676" width="1" height="1" alt=""><hr><p><a href="https://medium.com/carsguide-engineering/going-all-in-with-kubernetes-part-2-9d6af24d5676">Going all in with Kubernetes (Part 2)</a> was originally published in <a href="https://medium.com/carsguide-engineering">CAMS Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Going all in with Kubernetes (Part 1)]]></title>
            <link>https://medium.com/hackernoon/going-all-in-with-kubernetes-part-1-be0852a87ea2?source=rss-72aea1791bfe------2</link>
            <guid isPermaLink="false">https://medium.com/p/be0852a87ea2</guid>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[docker]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[going-all-with-kubernetes]]></category>
            <dc:creator><![CDATA[Richard Sands]]></dc:creator>
            <pubDate>Tue, 29 May 2018 06:37:01 GMT</pubDate>
            <atom:updated>2018-06-07T15:07:09.149Z</atom:updated>
            <content:encoded><![CDATA[<p>I use to think the dream would be when we (carsguide) roll out Docker containers on our AWS infrastructure and take baby steps to containerization, instead we went all in with Kubernetes — head first. On the face of it we had two choices; either take a risk and reap or do nothing and remain in the “safe zone”. I like to take risks but being honest, I wasn’t sure we were ready for Kubernetes. Turns out I was wrong to some degree.</p><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*Vu4VoJDRH2NO40DqwLb1gw.png" /></figure><p>Traditionally this is what happens in many software teams; engineers build things, package it up and hand over to ops who manage a pipeline, deploy and run it. (Our ops team built all of the current pipelines). The flow looks a bit like this;</p><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*fPo-CtgS4LXRVxnbI5L8Ng.jpeg" /><figcaption>Most teams are split</figcaption></figure><p>I am a big believer in giving engineers freedom to build how they want, deploy how they want. Engineers like ownership and it makes engineers much stronger if they know how to deploy, run and monitor what they are building.</p><p>Early 2018 a few of our tech team sat around a table and agreed to engage with a <a href="https://www.vibrato.com.au/">devops consultancy firm, Vibrato</a>. The initial agreement was to investigate possible CI/CD workflows and how to run our stack. Out of this engagement came a choice; go all in with Kubernetes or stay safe and use docker/ECS.</p><blockquote>Long story short we went all in with Kubernetes.</blockquote><p>A few key reasons;</p><ul><li>We are building a new micro service platform</li><li>Cost optimisation due to above reason</li><li>Engineers can build it from start to finish <em>alongside </em>operations, it is the future of engineering</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*ob1laG9FelkmmPO6DICCnQ.jpeg" /><figcaption>What we really want our teams to be</figcaption></figure><p>In a <a href="https://engineering.carsguide.com.au/monolith-applications-do-or-dont-fc1da6bf17b9">previous post</a> I talked about why micro services. In a nutshell; we are building a new platform. All of this platform is being built around micro services running on Docker containers. Given we will have many more services than we currently do we need a quick and easy way to deploy and scale each service without the overhead of configuring ALBs, SSL certificates, autoscaling groups, target groups, EC2 images, etc. Kubernetes does all of this for us using code defined in each service, it orchestrates the services for us and the networking to make it happen.</p><p>Think about it this way; an application may only ever need 10% CPU or 10% of the memory an EC2 has, that is <strong>90%</strong> wastage if you run a single application per EC2 like we do. To reduce costs we need to orchestrate many containers on a single EC2 (node). Kubernetes solves this, it does this by orchestrating the best possible utilisation of the compute nodes.</p><p>Our new Kubernetes staging cluster is running on 7 EC2s with over 95 pods (applications) running 200+ containers, load balanced with high availability. Our current staging takes over 50 EC2s (currently 126(!!) EC2s if you count all of our code bases). Kubernetes should save <strong>thousands </strong>of dollars per month.</p><p>Our current pipelines are complex, most of the engineers do not know how they work or were to start if they wanted to create a new pipeline. Kubernetes has let us define how the service should run from the code in each service repository.</p><p>If an engineer wants 5 pods for a service they set an autoscaler definition to 5 and push to the repository. A few minutes later they get 5 pods. If you want cron jobs running, again define them in the code, push and a few minutes later your cron jobs will run. Kubernetes allows this to happen, everything is in code.</p><p>Engineers are able to create a pipeline, setup service requirements, how it scales, what containers it needs and deploy to staging in a matter of hours. No longer will it take weeks.</p><blockquote>Services are defined in code that any engineer can follow without knowledge of the cloud provider. GCP, AWS, Azure — doesn’t matter, it is the same code that will do the same task no matter what it runs on.</blockquote><p>I am going to follow up this post with a few more on our journey. But the above gives you insights to the main reasons why we decided to try it. Some things I am aiming to cover;</p><ul><li>Tooling we use for CI/CD</li><li>Problems we have faced so far (and there have been a few)</li><li>Logging and alerting</li><li>You build it you own it</li></ul><p>Have you started a journey on Kubernetes? Always keen to hear other peoples views on their experiences.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=be0852a87ea2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/hackernoon/going-all-in-with-kubernetes-part-1-be0852a87ea2">Going all in with Kubernetes (Part 1)</a> was originally published in <a href="https://medium.com/hackernoon">HackerNoon.com</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>