Talk in English - UK at PHP North West 2017
View Slides: https://speakerdeck.com/gbtekkie/phpnw2017-smart-scalable-content-distribution
Short URL: https://joind.in/talk/7c16e
(QR-Code (opens in new window))
Separating data from functionality is one of the tricks to speedier applications, since it allows accessing the right content at the right time. Partitioning data cleverly presents several challenges that we've overcome for several projects, and it requires a combination of architecture, DevOps, development and testing skills that is pretty unique. Unless you have such a swiss-army-knife in your team (and are not afraid of the bus factor), you'd rather want to come listen to how we do it.
Comments
Comments are closed.
Very conceptual talk, aimed at high load/availability sites, would be interested to watch the talk again after do some further reading around some of the areas to understand it better.
Good talk on scalable content distribution. Really interesting to hear about separating the CMS backend & front-end delivery, something that seems a really good idea for many use cases. Nice content / CMS theory and a couple of excellent book recommendations I’ll check out (Beyond Legacy Code & Complete Web Monitoring).
High level talk explaining the architecture behind splitting content and your application. Found it was quite hard to follow at times, but I imagine it will make more sense with some further reading.
I might be the wrong person to review the talk, as we develop web apps rather than marketing/ecommerce. I think the overall implementation wasn't clear - if you disconnect the Wordpress content db from the front end, what renders the front end, i.e. is it just Wordpress somewhere else and how does that link to the database, or is it static rendering and how does that get carried out? There's really interesting concepts here though - the idea of replacing a dynamic CMS with edit-side CMS and static fast content has been well used at LADBible (they deploy to S3). I also think (again very much based on my background) that a talk about content distribution would have an easy win touching on CDNs, and sort of expected this from the abstract.
I found this talk extremely interesting, and covers a cost-effective solution to massively scaling an existing platform, without having to rearchitect the entire thing. Will certainly be easily achievable as frameworks move to be more headless.
There were a couple of challenges with the talk:
1. The description in the programme was really ambiguous - I was not sure what to expect when I arrived in the session (and it turned out that I'd misunderstood the purpose of the talk).
2. Throughout the talk, I was unsure what the actual solution was that was being proposed. I imagine to those that have worked with it, it will have seemed clear, but for me, I kept hearing about 'content' being moved into 'silos' but I was never clear what those 'silos' were. Some clear examples of this with some implementation details would have made it a very different talk.
It was literally right at the end in the questions when I understood what it was about - I could then see potential in the theory, so now have food for thought.