<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: bddicken</title><link>https://news.ycombinator.com/user?id=bddicken</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Wed, 18 Nov 2026 14:46:46 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=bddicken" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[The latency inside your database]]></title><description><![CDATA[
<p>Article URL: <a href="https://planetscale.com/blog/the-hidden-latency-inside-your-database">https://planetscale.com/blog/the-hidden-latency-inside-your-database</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=50405485">https://news.ycombinator.com/item?id=50405485</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Sat, 07 Nov 2026 15:44:07 +0000</pubDate><link>https://planetscale.com/blog/the-hidden-latency-inside-your-database</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=50405485</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50405485</guid></item><item><title><![CDATA[What today's software owes to Bronze Age clocks]]></title><description><![CDATA[
<p>Article URL: <a href="https://planetscale.com/blog/leaky-buckets-in-software-explained">https://planetscale.com/blog/leaky-buckets-in-software-explained</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49998804">https://news.ycombinator.com/item?id=49998804</a></p>
<p>Points: 5</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 07 Oct 2026 21:06:58 +0000</pubDate><link>https://planetscale.com/blog/leaky-buckets-in-software-explained</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=49998804</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49998804</guid></item><item><title><![CDATA[New comment by bddicken in "Tin: full-text search for Postgres"]]></title><description><![CDATA[
<p>THIS++</p>
]]></description><pubDate>Sat, 19 Sep 2026 18:45:03 +0000</pubDate><link>https://news.ycombinator.com/item?id=49769123</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=49769123</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49769123</guid></item><item><title><![CDATA[New comment by bddicken in "Massively Parallel Postgres Backups"]]></title><description><![CDATA[
<p>pgBaseBackup seems very cool. It's great that they got support to continue building. We just don't use it presently.</p>
]]></description><pubDate>Tue, 04 Aug 2026 03:23:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=49164072</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=49164072</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49164072</guid></item><item><title><![CDATA[New comment by bddicken in "Massively Parallel Postgres Backups"]]></title><description><![CDATA[
<p>Neki will support cross-shard ACID transactions with a combination of an external transaction coordinator and some changes to PostgreSQL itself to support this (either by engine modifications or extensions).<p>I expect as part of that, we'll allow users to leverage something similar to make sure that we don't get transaction tearing in backups.</p>
]]></description><pubDate>Mon, 03 Aug 2026 23:24:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=49162651</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=49162651</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49162651</guid></item><item><title><![CDATA[New comment by bddicken in "Massively Parallel Postgres Backups"]]></title><description><![CDATA[
<p>Exactly. When you have a router + query parser in front of the db nodes, the burden of managing multiple versions simultaneously can be taken out of the app and into the database layer.</p>
]]></description><pubDate>Mon, 03 Aug 2026 21:23:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=49161628</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=49161628</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49161628</guid></item><item><title><![CDATA[New comment by bddicken in "Massively Parallel Postgres Backups"]]></title><description><![CDATA[
<p>Broadly for Pg, minor version upgrades are straightforward as the data on disk is guaranteed to be compatible. All it takes is a restart or switchover to upgrade.<p>Major versions are more challenging for "vanilla" Postgres because that's not the case. Storage/catalog formats may change. There needs to be an explicit upgrade process for the data (eg, a database created with v18 won't work out of the box with v19).<p>The cool thing is, software like Neki (and Vitess for MySQL, which we maintain) have architectures that lend themselves to making this much more feasible. Because the actual database nodes sit behind a router + parser layer with Pg/MySQL compatibility, the upgrades can be done transparently to the user. We plan to write more about how this works in the future.</p>
]]></description><pubDate>Mon, 03 Aug 2026 21:13:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=49161510</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=49161510</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49161510</guid></item><item><title><![CDATA[New comment by bddicken in "Massively Parallel Postgres Backups"]]></title><description><![CDATA[
<p>What I emphasized in this article (author here) is scaling postgres backups. This works because there's already rock-solid systems built into postgres + surrounding tooling to build from.<p>The broader takeaway is the principle of, "how do I take something that doesn't scale on its own and make it so?" This applies to backups, compute, storage layers, proxies. It's why Neki and Vitess are so powerful for everything from small 1GB databases to petabytes.<p>Hanging around to answer questions, too :)</p>
]]></description><pubDate>Mon, 03 Aug 2026 19:28:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=49160316</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=49160316</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49160316</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>Author here, thank you.<p>Technically, they are using js + gsap + svg embedded i the article with iframes.<p>Process-wise, I drafted most of them as static images in excalidraw, passed the images along to cursor for a first draft, applied styling rules, and then did a bunch of fine-tuning.</p>
]]></description><pubDate>Thu, 16 Jul 2026 15:14:43 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935718</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935718</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935718</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>Very welcome</p>
]]></description><pubDate>Thu, 16 Jul 2026 15:12:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935677</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935677</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935677</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>Yep! <a href="https://vitess.io/docs/reference/features/distributed-transaction/" rel="nofollow">https://vitess.io/docs/reference/features/distributed-transa...</a></p>
]]></description><pubDate>Thu, 16 Jul 2026 15:12:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935674</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935674</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935674</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>We have tons of customers who do exactly this. It's great. Sharding is for customers who out grow this path.</p>
]]></description><pubDate>Thu, 16 Jul 2026 15:05:59 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935597</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935597</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935597</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>A lot of what you're referring to is dictated by the sharding strategy. Vitess and Neki both let you configure this via the VSchema / data topology (That's what I'm getting at here)<p><a href="https://planetscale.com/blog/making-768-servers-look-like-1#how-does-it-know" rel="nofollow">https://planetscale.com/blog/making-768-servers-look-like-1#...</a><p>This decision matters a ton. I drilled into this specific point on sharding in one of my articles from awhile back:<p><a href="https://planetscale.com/blog/database-sharding" rel="nofollow">https://planetscale.com/blog/database-sharding</a><p>This is something you face any time you shard data, no matter the system. Single-shard queries are preferred to cross-shard ones. We want 99% of what we do to be single shard, but occasionally cross-shard work is unavoidable.</p>
]]></description><pubDate>Thu, 16 Jul 2026 15:02:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935558</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935558</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935558</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>You should read through those articles.</p>
]]></description><pubDate>Thu, 16 Jul 2026 14:58:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935499</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935499</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935499</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>> So an extreme example is OpenAI needing 50 replicas, but we're doing five blades ... err, we're doing 768 servers because the need arose "pretty quickly"?<p>If you read the OpenAI article, you'll see that they actually used sharding to offload a bunch of work from their "1 primary 50 replicas setup"<p>>>> "To mitigate these limitations and reduce write pressure, we’ve migrated, and continue to migrate, shardable (i.e. workloads that can be horizontally partitioned), write-heavy workloads to sharded systems such as Azure Cosmos DB..."<p>The 768 servers and 1PB example is just one of many configurations. A business with 10TB may choose to go from a monolithic database to a 8-shard setup to improve backup times, eliminate single-point-of-failure, have more breathing room for scaling.</p>
]]></description><pubDate>Thu, 16 Jul 2026 14:55:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935467</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935467</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935467</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>Hey, author here.<p>Technically, they're powered by js + gsap + svg.<p>Process wise (for most of them) I sketched them out in advance in excalidraw for figure out layout, then passed these along to cursor to have it build out an initial draft from the image, then used some styling rules to get all the styles inline, then did a bunch of fine-tuning.</p>
]]></description><pubDate>Thu, 16 Jul 2026 14:48:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935370</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935370</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935370</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>Spreading requests out across hundreds, thousands, and in some cases even more is precisely what is done in the industry for big databases! Good examples:<p>cashapp: <a href="https://code.cash.app/planetscale-metal" rel="nofollow">https://code.cash.app/planetscale-metal</a>
github: <a href="https://github.blog/engineering/infrastructure/partitioning-githubs-relational-databases-scale/" rel="nofollow">https://github.blog/engineering/infrastructure/partitioning-...</a>
etsy: <a href="https://www.etsy.com/codeascraft/migrating-etsyas-database-sharding-to-vitess" rel="nofollow">https://www.etsy.com/codeascraft/migrating-etsyas-database-s...</a><p>These companies could not realistically operate off a single database server.</p>
]]></description><pubDate>Thu, 16 Jul 2026 14:42:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=48935292</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48935292</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48935292</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>Caching at all levels is key to good db performance (cpu cache <-> ram <-> disk). CPU cache optimization is not my area of expertise, but I did write another fun article on io devices and how it related to databsae perf:<p><a href="https://planetscale.com/blog/io-devices-and-latency" rel="nofollow">https://planetscale.com/blog/io-devices-and-latency</a><p>Network saturation, and just resource saturation broadly, is a huge reason to shard. If/When the network, cpu, or disk for a subset of shards becomes a bottleneck, add more shards.</p>
]]></description><pubDate>Wed, 15 Jul 2026 22:07:07 +0000</pubDate><link>https://news.ycombinator.com/item?id=48927716</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48927716</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48927716</guid></item><item><title><![CDATA[New comment by bddicken in "Making 768 servers look like 1"]]></title><description><![CDATA[
<p>Hey, I wrote this!<p>Sharding is cool and foundational to making the internet work. I'm around to answer Qs.</p>
]]></description><pubDate>Wed, 15 Jul 2026 21:17:14 +0000</pubDate><link>https://news.ycombinator.com/item?id=48927100</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48927100</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48927100</guid></item><item><title><![CDATA[How to not screw up a benchmark]]></title><description><![CDATA[
<p>Article URL: <a href="https://planetscale.com/blog/on-benchmarking">https://planetscale.com/blog/on-benchmarking</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=48095328">https://news.ycombinator.com/item?id=48095328</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 11 May 2026 14:17:39 +0000</pubDate><link>https://planetscale.com/blog/on-benchmarking</link><dc:creator>bddicken</dc:creator><comments>https://news.ycombinator.com/item?id=48095328</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=48095328</guid></item></channel></rss>