Unlocked Conference recap blog #451
Conversation
Signed-off-by: Crystal Pham <cpham@linuxfoundation.org>
|
@stockholmux @madolson any feedback on this one? |
|
This is the PR for issue #455 |
madolson
left a comment
madolson
left a comment
There was a problem hiding this comment.
I'm not really sure what the takeaway from this blog is. It seems like a high level recap, I think it would be more compelling to have a few deep take ways. This seems more like maybe a social post?
Thanks for the feedback! It’s very much a wide-net approach—high-level, focused on showcasing the event and generating momentum for the next unconference. We’ll definitely be sharing the blog on social, but the full context and details wouldn’t really fit into a single social post on their own. |
|
Hello, we have some deep-dive posts in the pipeline (one completed but on hold as the PR plan gets finalized). This "Event Recap" is targeted to highlight external validation of the project. By hosting on the Valkey blog, we show:
From a takeaways point of view, the "Recurring Patterns" section highlights a few of the specific production challenges discussed giving them context about why they matter to the broader community but we definitely plan deeper dives into some of the areas highlighted by presenters as we are allowed. If we don't feel this fits the Valkey blog directionally, we can move it to the Momento blog, but I believe the project benefits from having this type of "Proof of Life" on the official site. |
stockholmux
left a comment
stockholmux
left a comment
There was a problem hiding this comment.
- I don't know who the audience is supposed to be for this blog post. A very crisp target might make the post have a more solid direction.
- It says a lot of things without backing them up.
- Is there a call-to-action?
| The conference created a space for those lessons. It provided a room where engineers operating large, performance-sensitive systems could talk openly about what they’ve seen in production, what surprised them, and what they’d do differently next time. The audience for this conference was engineers already living with these systems every day and those searching for a new way to achieve their goals. | ||
|
|
||
| ## The Venue & Atmosphere | ||
| The inaugural Unlocked Conference was held on January 22 at Guildhouse in San Jose, California. It’s a venue built for competitive gaming and live events, and it was a perfect place to talk about high performance computing. Most importantly, it was a space that encouraged people to stay engaged between sessions instead of disappearing. |
There was a problem hiding this comment.
Drop the part about the venue. I don't think it adds much for the audience.
|
|
||
| These conversations were exactly the reason we put Unlocked together. The ability to talk through a problem in real time, sketch an architecture on a notepad, or compare approaches with someone who’s been there accelerates learning in a way that’s hard to replicate elsewhere. | ||
|
|
||
| ## By the Numbers |
There was a problem hiding this comment.
I don't think this is relevant to the audience.
|
|
||
| There was also a recurring focus on interaction effects. Meaning multiple systems behaving correctly in isolation while producing unexpected behavior together, problems that only become obvious after running real workloads in production. | ||
|
|
||
| Many sessions, like [Ignacio Alvarez’s](https://www.linkedin.com/in/ignacioalvarez79/) on moving from Memcached to Redis to Valkey, described personal experiences: what broke, how it was diagnosed, and what actually improved things. Audience questions reflected that same depth. They came from engineers who had clearly run into similar situations and were comparing notes and looking for solutions. |
There was a problem hiding this comment.
Can there be links to sessions?
There was a problem hiding this comment.
I will let Allen or Mike confirm - from my knowledge, they are still under review
| That framing set the tone for the rest of the day. Talks were technical, experience-driven, and aimed at people building and operating production systems at scale. | ||
|
|
||
| ## Recurring Patterns | ||
| Several ideas kept resurfacing across the talks and our conversations. Tail latency came up repeatedly as a leading indicator that makes a system feel stable or unpredictable. Several speakers walked through scenarios where a small change in the client’s request shape or request mechanics had a significant impact on the systems’ performance at scale. |
There was a problem hiding this comment.
If I were reading this, I'd love to be able to find out more about tail latency, performance at scale but you're telling, not showing.
| ## Recurring Patterns | ||
| Several ideas kept resurfacing across the talks and our conversations. Tail latency came up repeatedly as a leading indicator that makes a system feel stable or unpredictable. Several speakers walked through scenarios where a small change in the client’s request shape or request mechanics had a significant impact on the systems’ performance at scale. | ||
|
|
||
| There was also a recurring focus on interaction effects. Meaning multiple systems behaving correctly in isolation while producing unexpected behavior together, problems that only become obvious after running real workloads in production. |
| ## Hallway Conversations | ||
| Between sessions, the hallway track picked up right where the talks left off. Small groups formed around shared pain points: performance regressions, migration challenges, and edge cases that felt familiar across companies. More than once, conversations turned into quiet moments of recognition where someone described a problem and others nodded along because they’d seen it too. | ||
|
|
||
| One of the things we overheard was infrastructure leaders sharing that the event gave them the confidence to accelerate their own migrations to Valkey. They shared that watching industry leaders like Apple, Uber, and Snap solve hard performance problems with Valkey gave them the confidence that the project is ready. |
There was a problem hiding this comment.
'One of the things we overheard' feels creepy. Like you're eavesdropping. I think it's fine to state that people discussed something in the hallway track, but this framing doesn't come off as respectful.
|
|
||
| One of the things we overheard was infrastructure leaders sharing that the event gave them the confidence to accelerate their own migrations to Valkey. They shared that watching industry leaders like Apple, Uber, and Snap solve hard performance problems with Valkey gave them the confidence that the project is ready. | ||
|
|
||
| These conversations were exactly the reason we put Unlocked together. The ability to talk through a problem in real time, sketch an architecture on a notepad, or compare approaches with someone who’s been there accelerates learning in a way that’s hard to replicate elsewhere. |
There was a problem hiding this comment.
'The ability to talk through a problem in real time, sketch an architecture on a notepad, or compare approaches with someone who’s been there accelerates learning in a way that’s hard to replicate elsewhere.' is a good hook vs 'hallway converstaions'
| Companies represented: 23 | ||
| Sessions delivered: 15 | ||
|
|
||
| We will continue to grow the Unlocked Conference as a place for engineers to share lessons learned under real operating conditions. For those who weren’t able to attend in person, the sessions were recorded and will be available to watch soon. [Join us at the next Unlocked Conference](https://unlocked.gomomento.com/) in early May 2026! More information coming soon - subscribe to the [Valkey Community newsletter](https://hubs.la/Q03-LC830) to get the latest updates. |
There was a problem hiding this comment.
You're using 'we' a lot - it needs to be very crisp that the 'we' isn't the project (Momento? Momento + AWS?)
|
|
||
| The inaugural Unlocked Conference was a strong start, and it’s clear there’s real momentum behind continuing these conversations. | ||
|
|
||
| Thank you to everyone at [Momento](https://www.gomomento.com/) and [AWS](https://aws.amazon.com/) who hosted such a wonderful event. And special thanks to the lunch sponsor, [EloqData](https://www.eloqdata.com/) 💚 No newline at end of file |
There was a problem hiding this comment.
Feels weird: the authors come from momento... so you're thanking yourself?
| --- | ||
| title: Mike Callahan | ||
| extra: | ||
| photo: '/assets/media/authors/mikecallahan.jpeg' |
There was a problem hiding this comment.
this photo is not included in the PR so it's broken on the live preview of the site.
The audience is engineering leaders and decision making infrastructure architects. It shows those that are in a wait-and-see mode of migrating that others (Apple, AWS, Snap, Uber, etc) have taken the leap.
Generally it's a similar target and format as the https://valkey.io/blog/2025-year-end/ blog but with additional company mentions.
|
Alright. I would really focus the content on that audience. Stuff like descriptions of the venue and number of attendees would be tough matches for that profile. A good exercise would be to imagine yourself as one of these folks and look at each paragraph and ask "Would I care?"
That post was community-oriented. I'm guessing it performed very poorly for audience you're targeting. For your target, my guess would be that engineering leaders and architects will want technical but high-level descriptions of what happened and a way to find out more. They're probably thinking "was this conference relevant to my current set of challenges"? So, if you can answer those type of questions about this one, it's a great opening to push the May conference at the end of the post. Regarding the CTA, I think it would be more effective at the end of the blog post as it gives readers a next step. If they're going to click on either of those items it's a odd flow to have to come back and read the rest of the post. |
5 participants
Description
Unlocked Conference recap blog by the Momento team
Issues Resolved
Check List
--signoffBy submitting this pull request, I confirm that my contribution is made under the terms of the BSD-3-Clause License.