#yugabyte — Public Fediverse posts

#DOAG2025 hat nur noch wenige Aussteller :-( Lichtblick: #Yugabyte

The best time to start blogging is to not stop blogging in the first place. The next best time is today.

Here's a syndication of a post from the ComplyAdvantage Technology Blog to kicks things off.

https://mark.allanson.org/blog/dss-2024-ca-lessons/

#Yugabyte #YugabyteDB #DistributedSQL #ComplyAdvantage

Container Attached Storage (CAS) and Container Storage Interface (CSI) are the most important terms when it comes to Kubernetes and stateful workloads like databases (such as #postgresql, #yugabyte, and others). We explain them - just for you!

https://www.simplyblock.io/post/container-attached-storage-and-container-storage-interface-the-building-blocks-of-kubernetes-storage

Guest @FranckPachot from #yugabyte joins our very own @noctarius2k in this episode of the weekly, 20 min #CloudCommute #podcast, talking about #distributedsql, #postgresql , #vectordatabases, and more. Tune in!

The 🎙️ is available on Spotify, iTunes, Pandora, Amazon Music, and more.

🎥👉 https://youtu.be/1EAKqwcP2SY

#vectordatabase #vectorsearch #vectordb #database #databases #postgres #postgressql

Using #PostgreSQL RLS for multi-tenancy SaaS is quite common, but some users want to access a list of tenants. Here's how to do it efficiently (and leveraging loose index scan in #Yugabyte​DB)
https://dev.to/yugabyte/postgresql-row-level-security-with-an-array-of-tenants-2136

Batching remote reads/writes is how #Yugabyte​DB can provide elasticity with high performance. An example of Foreign Key checking:
https://dev.to/yugabyte/foreign-keys-on-distributed-sql-dont-worry-it-scales-2g0g

An example of read restart, which is how Oracle Database or #Yugabyte​DB avoids serialization errors in Read Committed, without inconsistent read time:

https://dev.to/yugabyte/dirty-writes-insert-on-conflict-do-update-read-committed-isolation-level-and-lost-update-13hk

What? There's no Inverted Join in #Yugabyte​DB?
Of course not. That's not how SQL works.
If you want a Nested Loop to an Inverted Index, do like in PostgreSQL: declare a GIN index and let the DB choose the join method:
https://dev.to/yugabyte/inner-inverted-join-in-yugabytedb-postgresql-g22

If you want to be sure that your SELECT COUNT(*) is optimal in #Yugabyte​DB, check the execution plan. There are many optimisations where it uses PostgreSQL with additional pushdowns to reduce the latency
https://dev.to/yugabyte/fast-select-count-in-yugabytedb-464h

The benefits of hash-sharding (to avoid hotspots) and range-sharding (to reduce reads) can be combined in #Yugabyte​DB, thanks to LSM-Tree scan optimization and PostgreSQL features: https://dev.to/yugabyte/scalable-range-sharding-with-yugabytedb-1o51

#Yugabyte​DB uses #PostgreSQL query layer plugged on top of a cloud-native transactional distributed store. In addition to scaling out, it's the occasion to address some PostgreSQL problems (bloat, vacuum, connection processes, HA, upgrades...) https://www.yugabyte.com/blog/improve-postgresql/

🔥Curious if @Citusdata is Distributed SQL like #Yugabyte​DB?💡Nope! Citus rocks sharding over PostgreSQL databases, but it's not a SQL database on its own. Remember, cross-shard reads through Citus coordinator aren't ACID but eventually consistent https://dev.to/yugabyte/citus-is-not-acid-but-eventually-consistent-3711

Counters in SQL 🤓 I tested #Yugabyte​DB compaction with this use-case
👉 no vacuum, no bloat 🎶

https://dev.to/yugabyte/in-memory-counters-with-yugabytedb-2p54

You can easily build and install many PostgreSQL extensions on #Yugabyte​DB - here is an example with pg-hostname:
https://dev.to/yugabyte/pg-hostname-on-yugabytedb-compiled-and-installed-directly-on-the-server-15id

New blog post: #SQLPage. If you💗SQL you can build your web app presentation layer with SQL declarations. What if the DB is unavailable? Just use #Yugabyte​DB: It is Distributed PostgreSQL, with no downtime during failures or upgrades 🚀https://dev.to/yugabyte/sql-for-low-code-applications-sqlpage-lmk

Because I've seen some silly benchmarks comparing 🍏to 🍊, I had a look at #hammerdb on #Yugabyte​DB. With #PostgreSQL-compatible databases, you can know what you run

https://dev.to/yugabyte/setup-ssl-for-ysql-in-yugabyte-manually-7hd #yugabyte

"LSM-tree and block storage" by @fritshoogland #DEVCommunity https://dev.to/yugabyte/lsm-tree-and-block-storage-29ek #yugabyte #yugabyteDB

@fritshoogland that is very cool, the fact that you are reusing codebase here. That alone tells me a lot!

I am very interested on this, yet I have no applicable use. Hope you don't mind if I ask you a bunch of questions.

Or... if you want to direct me to read something in particular, that's good too 👍

I basically want to know "what are we giving up" when using #Yugabyte instead of #PostgreSQL. (cons on single server and/or low count of servers)

@fritshoogland I wonder, how close is #Yugabyte to #PostgreSQL? Is the query planning as good? does it support the same level of SQL? How well it performs in single-cluster vs postgreSQL?

I would be interested on seeing benchmarks of complex queries with >10GiB databases to see how much performance would be lost compared to #PostgreSQL

If it's not much, Yugabyte might be interesting for DBs that you might to run in cloud, even for regular data sizes, just in case it becomes too big later on.