Sure the dbaas might have superb autoscing and reliability. It may or may not have good OpenTofu/terraform providers available, or other custom niche deployment tools one would need to pick up & adapt.
Same questions as roughing it hacking together pg yourself, except now you have no power. Is observability going to be up to snuff? When performance problems come, how are you going to feel when you discover that analyst techniques a and b work fine but c and d can't be done because you are on managed service that doesn't actually give you full access to the db?
Costs can also add up, and are hard to predict. If you have a managed service, you'll be using a lot more network, which often has some cost (also some latency too). This isn't for everyone, but Kube makes it easy to start with anything (ex: RPi) and scale out. Works with hosted hardware with some local SSD, or I can get a 16-core 7950X and a bunch of gen 5 nvme drives and a fraction of a terabyte of ram for <$2000 and that'll scale me to godlike levels. Whereas I'll keep paying for something hosted.
I think it's so so ideal to have a reliably capable universal autonomic computing later underfoot. Most of the backend has had special specific answers to each question, each problem. We have been disjointed. Kubernetes provides a platform applicable to a vast range of workloads, where yes you need to tune and build different clusters for different workloads, but where the skills transfer much more readily. Everyone in Kubernetes knows how to ship resources definitions/manifests, and that base knowledge is all you need to learn to start consuming services - any services - on Kubernetes. Skill transfer is immediate. Having universal patterns (the API server), backed by autonomic operators making & keeping these resources real: it's so much better than the thousand different paths road of technology we've been on until now."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Self-hosting a high-availability Postgres cluster on Kubernetes"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://ryan-schachte.com/blog/ha_postgres_zolando/"}},"_tags":["comment","author_jauntywundrkind","story_38843589"],"author":"jauntywundrkind","comment_text":"These are each specific DB-as-a-services. Instead of sinking my time into learning something universal & competent & capable at many things (Kubernetes) now I'm investing in learning & using a niche service, that I won't have fine control over & whose limits might only be clear down the road. That's a risk.
Sure the dbaas might have superb autoscing and reliability. It may or may not have good OpenTofu/terraform providers available, or other custom niche deployment tools one would need to pick up & adapt.
Same questions as roughing it hacking together pg yourself, except now you have no power. Is observability going to be up to snuff? When performance problems come, how are you going to feel when you discover that analyst techniques a and b work fine but c and d can't be done because you are on managed service that doesn't actually give you full access to the db?
Costs can also add up, and are hard to predict. If you have a managed service, you'll be using a lot more network, which often has some cost (also some latency too). This isn't for everyone, but Kube makes it easy to start with anything (ex: RPi) and scale out. Works with hosted hardware with some local SSD, or I can get a 16-core 7950X and a bunch of gen 5 nvme drives and a fraction of a terabyte of ram for <$2000 and that'll scale me to godlike levels. Whereas I'll keep paying for something hosted.
I think it's so so ideal to have a reliably capable universal autonomic computing later underfoot. Most of the backend has had special specific answers to each question, each problem. We have been disjointed. Kubernetes provides a platform applicable to a vast range of workloads, where yes you need to tune and build different clusters for different workloads, but where the skills transfer much more readily. Everyone in Kubernetes knows how to ship resources definitions/manifests, and that base knowledge is all you need to learn to start consuming services - any services - on Kubernetes. Skill transfer is immediate. Having universal patterns (the API server), backed by autonomic operators making & keeping these resources real: it's so much better than the thousand different paths road of technology we've been on until now.","created_at":"2024-01-03T17:46:51Z","created_at_i":1704304011,"objectID":"38857073","parent_id":38855083,"story_id":38843589,"story_title":"Self-hosting a high-availability Postgres cluster on Kubernetes","story_url":"https://ryan-schachte.com/blog/ha_postgres_zolando/","updated_at":"2024-09-20T16:03:53Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rektide"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Unmentioned so far:
* Form History Control (rip Lazarus Forms Recovery, Seth plz come back & open source it). https://stephanmahieu.github.io/fhc-home/Manual/manual/
* Multi-Account Containers in Firefox. Absolutely critical to mixing work/personal computing. I used to use profiles to split stuff up but harder to maintain. (oops already mentioned)
* React Developer Tools. Apollo Developer Tools. Instrumental to my development work.
* ViolentMonkey. Before Chrome finally blows up freedom & expression on the web: ViolentMonkey is a pretty good userscripting tool. https://violentmonkey.github.io/
* Open in Sci-hub. I came very close to weaping out loud when I found that these papers I have forever clicked on & wanted are easily within reach. I now have access to humankind's wisdom. https://roiarthurb.github.io/Side-Auto_Sci-Hub/
* Hypothesis. The greatest most pro-web tool humanity has ever imagined, by orders of magnitude. https://web.hypothes.is/start/
* Web Scrobbler. I love knowing what I listened to. Works great with usual suspects. So so results with individual online fm radio stations. https://web-scrobbler.com/
I really want to get good at a vim<->browser extension. Ideally just using a real vim session. I use Dark Reader all the time & it's amazing, just fantastic. I want a replacement for Share-a-holic that I trust, that let's me bookmark/send-to across multiple services (pocket, pinboard primarily); I don't trust share-a-holic. "Switch to audible tab" is useful in critical moments, which thankfully are not often. Extensions I'd like to find & get good at: semantic web extractors/breadcrumbers."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: What browser extensions are a must-have in 2021?"}},"_tags":["comment","author_rektide","story_29314560"],"author":"rektide","comment_text":"Unmentioned so far:
* Form History Control (rip Lazarus Forms Recovery, Seth plz come back & open source it). https://stephanmahieu.github.io/fhc-home/Manual/manual/
* Multi-Account Containers in Firefox. Absolutely critical to mixing work/personal computing. I used to use profiles to split stuff up but harder to maintain. (oops already mentioned)
* React Developer Tools. Apollo Developer Tools. Instrumental to my development work.
* ViolentMonkey. Before Chrome finally blows up freedom & expression on the web: ViolentMonkey is a pretty good userscripting tool. https://violentmonkey.github.io/
* Open in Sci-hub. I came very close to weaping out loud when I found that these papers I have forever clicked on & wanted are easily within reach. I now have access to humankind's wisdom. https://roiarthurb.github.io/Side-Auto_Sci-Hub/
* Hypothesis. The greatest most pro-web tool humanity has ever imagined, by orders of magnitude. https://web.hypothes.is/start/
* Web Scrobbler. I love knowing what I listened to. Works great with usual suspects. So so results with individual online fm radio stations. https://web-scrobbler.com/
I really want to get good at a vim<->browser extension. Ideally just using a real vim session. I use Dark Reader all the time & it's amazing, just fantastic. I want a replacement for Share-a-holic that I trust, that let's me bookmark/send-to across multiple services (pocket, pinboard primarily); I don't trust share-a-holic. "Switch to audible tab" is useful in critical moments, which thankfully are not often. Extensions I'd like to find & get good at: semantic web extractors/breadcrumbers.","created_at":"2021-11-23T05:30:50Z","created_at_i":1637645450,"objectID":"29314897","parent_id":29314560,"story_id":29314560,"story_title":"Ask HN: What browser extensions are a must-have in 2021?","updated_at":"2024-09-20T09:48:51Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"theahura"},"title":{"matchLevel":"none","matchedWords":[],"value":"AI Agent does research, create full-length paper that passed blinded peer review"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://www.autoscience.ai/blog/ai-research-automation"}},"_tags":["story","author_theahura","story_43475289"],"author":"theahura","children":[43487638],"created_at":"2025-03-25T20:03:35Z","created_at_i":1742933015,"num_comments":1,"objectID":"43475289","points":2,"story_id":43475289,"title":"AI Agent does research, create full-length paper that passed blinded peer review","updated_at":"2025-03-26T21:55:23Z","url":"https://www.autoscience.ai/blog/ai-research-automation"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"frozenseven"},"title":{"matchLevel":"none","matchedWords":[],"value":"Carl: The First AI System to Produce Academically Peer-Reviewed Research"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://www.autoscience.ai/blog/meet-carl-the-first-ai-system-to-produce-academically-peer-reviewed-research"}},"_tags":["story","author_frozenseven","story_43403309"],"author":"frozenseven","created_at":"2025-03-18T18:52:07Z","created_at_i":1742323927,"num_comments":0,"objectID":"43403309","points":1,"story_id":43403309,"title":"Carl: The First AI System to Produce Academically Peer-Reviewed Research","updated_at":"2025-03-18T18:56:20Z","url":"https://www.autoscience.ai/blog/meet-carl-the-first-ai-system-to-produce-academically-peer-reviewed-research"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"fullung"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"When AWS Autoscale Doesn\u2019t"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://segment.com/blog/when-aws-autoscale-doesn-t/"}},"_tags":["story","author_fullung","story_19106767"],"author":"fullung","children":[19107331,19107359,19107423,19107668,19107679,19107855,19107910,19107970,19108615,19108833,19108842,19109542,19110233,19110671,19111675,19111815,19112453,19116502,19119411],"created_at":"2019-02-07T17:37:20Z","created_at_i":1549561040,"num_comments":92,"objectID":"19106767","points":203,"story_id":19106767,"title":"When AWS Autoscale Doesn\u2019t","updated_at":"2024-09-20T03:39:02Z","url":"https://segment.com/blog/when-aws-autoscale-doesn-t/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"drmorr"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Open-sourcing Clusterman, a cluster autoscaler for Kubernetes and Mesos"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://engineeringblog.yelp.com/2019/11/open-source-clusterman.html"}},"_tags":["story","author_drmorr","story_21546534"],"author":"drmorr","children":[21547186,21547983,21548094,21548825,21549005,21551130],"created_at":"2019-11-15T17:08:28Z","created_at_i":1573837708,"num_comments":44,"objectID":"21546534","points":109,"story_id":21546534,"title":"Open-sourcing Clusterman, a cluster autoscaler for Kubernetes and Mesos","updated_at":"2024-09-20T05:09:43Z","url":"https://engineeringblog.yelp.com/2019/11/open-source-clusterman.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ngaut"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Going Fast and Cheap: How We Made Anna Autoscale"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://rise.cs.berkeley.edu/blog/going-fast-and-cheap-how-we-made-anna-autoscale/"}},"_tags":["story","author_ngaut","story_17944956"],"author":"ngaut","children":[17950289,17950388,17950798,17951013,17951876,17953114],"created_at":"2018-09-09T08:30:11Z","created_at_i":1536481811,"num_comments":12,"objectID":"17944956","points":98,"story_id":17944956,"title":"Going Fast and Cheap: How We Made Anna Autoscale","updated_at":"2024-09-20T02:59:28Z","url":"https://rise.cs.berkeley.edu/blog/going-fast-and-cheap-how-we-made-anna-autoscale/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"alien_"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"My take at making AWS EC2 cheaper by automating SPOT instances with AutoScaling"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://mcristi.wordpress.com/2016/04/21/my-approach-at-making-aws-ec2-affordable-automatic-replacement-of-autoscaling-nodes-with-equivalent-spot-instances/"}},"_tags":["story","author_alien_","story_11539048"],"author":"alien_","children":[11539240,11539912,11540236,11540318,11542609,11542615],"created_at":"2016-04-21T01:32:10Z","created_at_i":1461202330,"num_comments":33,"objectID":"11539048","points":83,"story_id":11539048,"title":"My take at making AWS EC2 cheaper by automating SPOT instances with AutoScaling","updated_at":"2024-09-19T23:06:51Z","url":"https://mcristi.wordpress.com/2016/04/21/my-approach-at-making-aws-ec2-affordable-automatic-replacement-of-autoscaling-nodes-with-equivalent-spot-instances/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"AndrewDucker"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Autoscaling, Welcome to Google Compute Engine"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"http://googlecloudplatform.blogspot.com/2014/11/autoscaling-welcome-to-google-compute.html"}},"_tags":["story","author_AndrewDucker","story_8618818"],"author":"AndrewDucker","children":[8618931,8619002,8619075,8619189,8619676,8619710,8622518],"created_at":"2014-11-17T16:17:26Z","created_at_i":1416241046,"num_comments":31,"objectID":"8618818","points":68,"story_id":8618818,"title":"Autoscaling, Welcome to Google Compute Engine","updated_at":"2024-09-19T21:15:41Z","url":"http://googlecloudplatform.blogspot.com/2014/11/autoscaling-welcome-to-google-compute.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"mqzaidi"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Autoscaling on AWS"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"http://qzaidi.github.io/2013/06/21/autoscaling-with-aws/"}},"_tags":["story","author_mqzaidi","story_5916524"],"author":"mqzaidi","children":[5916603,5916624,5916682,5916772,5916821,5917023,5917073,5917297],"created_at":"2013-06-21T04:32:26Z","created_at_i":1371789146,"num_comments":28,"objectID":"5916524","points":68,"story_id":5916524,"title":"Autoscaling on AWS","updated_at":"2023-09-06T21:47:49Z","url":"http://qzaidi.github.io/2013/06/21/autoscaling-with-aws/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dschwartz88"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Show HN: Dynosaur, An autoscaler for Heroku using Google Analytics Live"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"http://engineering.harrys.com/2014/01/02/dynosaur-a-heroku-autoscaler.html"}},"_tags":["story","author_dschwartz88","story_7102723","show_hn"],"author":"dschwartz88","children":[7102744,7104487,7104653,7104942,7105185],"created_at":"2014-01-22T14:42:07Z","created_at_i":1390401727,"num_comments":23,"objectID":"7102723","points":66,"story_id":7102723,"story_text":"","title":"Show HN: Dynosaur, An autoscaler for Heroku using Google Analytics Live","updated_at":"2024-09-19T20:19:37Z","url":"http://engineering.harrys.com/2014/01/02/dynosaur-a-heroku-autoscaler.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"someofmyparts"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Show HN: kproximate \u2013 A Kubernetes node autoscaler for Proxmox"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/lupinelab/kproximate"}},"_tags":["story","author_someofmyparts","story_37713821","show_hn"],"author":"someofmyparts","children":[37715190,37716062,37716382,37717254],"created_at":"2023-09-30T09:10:11Z","created_at_i":1696065011,"num_comments":27,"objectID":"37713821","points":60,"story_id":37713821,"title":"Show HN: kproximate \u2013 A Kubernetes node autoscaler for Proxmox","updated_at":"2025-11-14T02:12:06Z","url":"https://github.com/lupinelab/kproximate"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"wanshao"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Show HN: AutoMQ - A Cost-Effective Kafka Distro That Can Autoscale in Seconds"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/AutoMQ/automq"}},"_tags":["story","author_wanshao","story_39966114","show_hn"],"author":"wanshao","children":[39966440,39966450,39966505,39966518,39966643,39966665,39966694,39966916,39966939,39968005,39968763,39968800],"created_at":"2024-04-08T04:05:15Z","created_at_i":1712549115,"num_comments":23,"objectID":"39966114","points":58,"story_id":39966114,"title":"Show HN: AutoMQ - A Cost-Effective Kafka Distro That Can Autoscale in Seconds","updated_at":"2026-04-10T03:16:04Z","url":"https://github.com/AutoMQ/automq"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"sytse"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"GitLab Runner 1.1 with Autoscaling"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://about.gitlab.com/2016/03/29/gitlab-runner-1-1-released/?"}},"_tags":["story","author_sytse","story_11382322"],"author":"sytse","children":[11382369,11383710,11384710],"created_at":"2016-03-29T15:48:51Z","created_at_i":1459266531,"num_comments":12,"objectID":"11382322","points":56,"story_id":11382322,"title":"GitLab Runner 1.1 with Autoscaling","updated_at":"2024-09-19T23:03:12Z","url":"https://about.gitlab.com/2016/03/29/gitlab-runner-1-1-released/?"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"innovate"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Autoscale Kubernetes workloads on any cloud using any event"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://kedify.io/resources/blog/kedify-keda-powered-public-beta-launch-announcement/"}},"_tags":["story","author_innovate","story_40213365"],"author":"innovate","children":[40213366,40215117,40215187,40215253,40216743,40225855],"created_at":"2024-04-30T17:10:11Z","created_at_i":1714497011,"num_comments":44,"objectID":"40213365","points":55,"story_id":40213365,"title":"Autoscale Kubernetes workloads on any cloud using any event","updated_at":"2024-10-02T04:33:13Z","url":"https://kedify.io/resources/blog/kedify-keda-powered-public-beta-launch-announcement/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"otterley"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Making Facebook\u2019s software infrastructure more energy efficient with Autoscale"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://code.facebook.com/posts/816473015039157/making-facebook-s-software-infrastructure-more-energy-efficient-with-autoscale/"}},"_tags":["story","author_otterley","story_8153798"],"author":"otterley","children":[8154180,8155980,8156157],"created_at":"2014-08-08T16:54:50Z","created_at_i":1407516890,"num_comments":26,"objectID":"8153798","points":55,"story_id":8153798,"story_text":"","title":"Making Facebook\u2019s software infrastructure more energy efficient with Autoscale","updated_at":"2023-09-06T22:08:58Z","url":"https://code.facebook.com/posts/816473015039157/making-facebook-s-software-infrastructure-more-energy-efficient-with-autoscale/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jonemi"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"GitHub, autoscaling, and the component substitution fallacy"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://surfingcomplexity.blog/2026/08/19/github-autoscaling-and-the-component-substitution-fallacy/"}},"_tags":["story","author_jonemi","story_49379026"],"author":"jonemi","children":[49380273],"created_at":"2026-08-20T19:28:18Z","created_at_i":1787254098,"num_comments":0,"objectID":"49379026","points":54,"story_id":49379026,"title":"GitHub, autoscaling, and the component substitution fallacy","updated_at":"2026-08-23T05:29:45Z","url":"https://surfingcomplexity.blog/2026/08/19/github-autoscaling-and-the-component-substitution-fallacy/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tzury"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"Compute Engine Predictive Autoscaling"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["autosci"],"value":"https://cloud.google.com/blog/products/compute/introducing-compute-engine-predictive-autoscaling"}},"_tags":["story","author_tzury","story_27719071"],"author":"tzury","children":[27719853,27719900,27719954,27720631,27721680,27721827],"created_at":"2021-07-03T03:14:09Z","created_at_i":1625282049,"num_comments":16,"objectID":"27719071","points":53,"story_id":27719071,"title":"Compute Engine Predictive Autoscaling","updated_at":"2024-09-20T08:51:56Z","url":"https://cloud.google.com/blog/products/compute/introducing-compute-engine-predictive-autoscaling"}],"hitsPerPage":20,"nbHits":110578,"nbPages":50,"page":0,"params":"query=AutoSci&advancedSyntax=true&analyticsTags=backend","processingTimeMS":19,"processingTimingsMS":{"_request":{"queue":9,"roundTrip":18},"fetch":{"query":6,"scanning":11,"total":18},"total":19},"query":"AutoSci","serverTimeMS":29}