Madelyn Olson

Some things kiro did:
1. Got frustrated at TCL, wrote it's own TCL framework
2. Wrote custom scripts to run valgrind and then wrote a readme for them
3. Wrote a brand new workflow.yaml file to include *JUST* the new test

Some things kiro did not do:
1. Actually notify the bio queue on shutdown

I do love kiro, but sometimes it feels like a toddler. I asked it make a small update, and then 15 minutes later:

> [bio_safe_shutdown 6d3ee79bd] Add safe shutdown mechanism for BIO threads
61 files changed, 42479 insertions(+), 5 deletions(-)
> Perfect! The implementation is complete.

The fact the url is "contact-sales/claude-for-oss" is really turning me off from recommending it to folks, I'm sure it's six months and then sales calls.

I wonder if generative AI is going to kill Amazon's writing culture. A doc that would have normally taken 1-2 weeks of thoughtful revision, can now be written by the frontier models with some MCPs to collect the data. The point of writing was to bring clarity. It's so easy to skip to the end now.

[NEW] Proposal: RESPB Binary Protocol for Valkey · Issue #2992 · valkey-io/valkey TL;DR RESPB is a binary wire protocol that replaces RESP text parsing with fixed-size opcodes and length-prefixed payloads. Working implementation shows 2.3x GET throughput, 46% lower SET latency, ...

We are considering a RESP4 (or maybe an extension to RESP3), github.com/valkey-io/va.... Specifically to better address multiplexing and improve performance. Part of the reason we're trying to have our own ecosystem.

A goal of Valkey is to mostly be a superset of the Redis APIs, but we also don't want to be constrained by decisions that Redis makes. We added new observability APIs in Valkey, and without client support it's harder for end users to adopt.

It's AWS funding most of it, it's not a secret. We actually started it before the fork, to try to have a more unified interface across clients. We've continued working on them after the fork.

🎉 Don’t miss out! The Unlocked Conference hits Jan 22 in San Jose, CA! #Valkey sessions you’ll love:

- Valkey in production
- Performance improvements in Valkey 9.0

🔥 Special deal: Register for just $99 with code LF99 (normally $200)!

✨ Spots are limited unlocked.gomomento.com/events/unloc...

Unlocked Conference 2026 A conference built for performance engineers who want to scale Valkey and push low-latency architectures to their limits.

I love small intimate conferences to spark creatively. Which is why I'm co-hosting a large scale performance focused conference in San Jose early next year with Momento. We have a lot of cool speakers signed up, come join if you're interested.

You can register at unlocked.gomomento.com

Releases · valkey-io/valkey A flexible distributed key-value database that is optimized for caching and other realtime workloads. - valkey-io/valkey

Last Friday, #Valkey released fixes for 7.2, 8.0, and 8.1 to address some Lua vulnerabilities that may allow RCE from an authenticated user. It's a good time to remind folks to always lock down your Valkey instances, even in secure environments. See github.com/valkey-io/va... for all the details)

YouTube video by Valkey Demystifying Valkey Clustering: Architecture, Fault Tolerance, and Scalability - Harkrishn Patro

I really enjoyed a talk by Harkrishn Patro, which dives into how the clustering system works in Valkey www.youtube.com/watch?v=P6Cb....

Keyspace 2025 - YouTube

If you weren't able to attend the recent Valkey event in Amsterdam, maybe you missed the Keyspace notification, you can see all the talks here www.youtube.com/playlist?lis... with slides posted valkey.io/events/keysp.... Check it out!

Today we announce DocumentDB has joined the LF with support from Microsoft, AWS and many more. Check it out.

www.prnewswire.com/news-release...

That's basically correct. It started out of customer conversations using ElastiCache without really understanding the nuanced durability/consistency tradeoffs. Once we explained it to them, many of them still chose to keep yolo'ing it though.

For cluster mode, we documented the procedure here for Valkey valkey.io/topics/clust..., but nothing has changed with respect to Redis AFAIK here.

I see some mention of using `CLIENT PAUSE`, but there is a safer variant in `FAILOVER` redis.io/docs/latest/... with standalone distributions. It will automatically pause incoming writes, wait for a replica to catch up, then orchestrate the failover.

Awesome! A fun fact about Valkey, is that the name originally came from ValkyrieDB. We change it because only about half of the folks on my team were able to spell Valkyrie.

A lot of people like `valkey-extended`. We are noodling between that valkey-bundle now. Thanks for the input!

So the #Valkey project is working on a packaged version of Valkey along with popular extensions (like LDAP authentication and vector similarity search). The plan is to call it `valkey-extensions`, but that the name might imply it's just extensions and not the core. Folks have better ideas/thoughts?

It's a bit last minute, but if anyone is interested in a live stream about tuning Valkey to run efficiently on multiple cores, we'll be hosting it in a couple of hours www.linkedin.com/events/73360.... There will be a recording as well.

A year on, Valkey charts path to v9 after break from Redis Interview: Fork focuses on stability and inclusion as it preps for more ambitious changes

Redis relicensed (under AGPL, which many cannot use). Valkey project, a Redis fork and still under an #opensource BSD license, continues to thrive in adoption and community at @linuxfoundation.org. Read more about what's next for Valkey:

www.theregister.com/2025/05/15/a... @valkeyio.bsky.social

Valkey Turns One: How the Community Fork Left Redis in the Dust - Momento Valkey is not only thriving, but now outperforming Redis 8.0 in real world benchmarks.

Just read a great blog by a friend about how to properly measure the performance of Valkey and Redis. When evaluating high performance systems, there are a lot of parameters which can affect the total throughput and it's easy to come up with "synthetic numbers". www.gomomento.com/blog/valkey-...

Reliable Cross-Cloud Kubernetes on Spot Instances with Architect Live migrate entire Kubernetes clusters — stateful workloads included — between AWS, Azure, GCP and bare metal with zero downtime. With Architect and PVM you can cut cloud costs by up to 90% and end v...

What if we could pick up a k8s cluster and move it w/out service disruption?

We demo’d this at Kubecon NA 2024 by moving a @kubernetes.io cluster running Valkey from AWS to Azure and then GCP all without disruption! this is pure magic from @felicitas.pojtinger.com .

loopholelabs.io/blog/zero-do...

Release 8.1.1 · valkey-io/valkey Upgrade urgency SECURITY: This release includes security fixes we recommend you apply as soon as possible. Security fixes (CVE-2025-21605) Limit output buffer for unauthenticated clients (#1994) ...

We have three new releases of #Valkey which include security fixes. Please consider upgrading if you have a publicly accessible Valkey instance or apply a mitigation to the CVE.
- github.com/valkey-io/va...
- github.com/valkey-io/va...
- github.com/valkey-io/va...

Photo of Madelyn Olson on stage at Monki Gras in front of a slide showing the quote “"We optimize for joy. We believe writing code is a lot of hard work, and the only way it can be worth is by enjoying it. When there is no longer joy in writing code, the best thing to do is stop. To prevent this, we'll avoid taking paths that will make Redis less of a joy to develop." by antirez

It’s @reconditerose.bsky.social quoting antirez at #MonkiGras on the importance of optimising for joy!

"[..] We believe writing code is a lot of hard work, and the only way it can be worth is by enjoying it. When there is no longer joy in writing code, the best thing to do is stop. [..]”

Maybe its time to try valkey.io, the modern open source alternative :)

Haha, oops. I'm a contributor to Valkey, so I've worked on it :). I know many obscure things about the engine.

I'm personally more excited about the memory efficiency stuff Valkey built though, you can read about it here: valkey.io/blog/valkey-...

I have too many alerts on mentions to Redis!

Single threaded, both Redis and Valkey are still pretty close. From my testing Valkey is faster but it is a bit use case specific. Multi-threaded there is a bigger lead for Valkey, but Redis is still working on their architecture, so things might change

It uses an approximate LRU. Each key has 3 bytes of LRU information, and when a key needs to get evicted it samples a few random items and picks the one with the earliest LRU timestamp. It saves 13 bytes compared to a linked list. You can read more information here: valkey.io/topics/lru-c....