There we were, at the second day of the rampconf event in Budapest. All I can say, well done organizers!
On the second day, there were fewer presentations but all of them was very interesting. We started the day with a tradition, having breakfast with other participants, for some this was a great opportunity to network a little, meet people from other companies, make some connections.
Around ten in the morning the first presentation was started, Scaling reddit from 1 million to 1 billion, pitfalls and lessons by Jeremy Edberg. I consider his presentation a really nice, well thought, well constructed, “multi-tier” presentation, because he broken up the concept of scalability and tried to present how does each part of a company/service affect/is affected by scalability or scalability issues. He talked about hardware, software, errors made, lessons learned and solutions. He underlined, that sometimes the community was the one who helped them out in problematic scenarios.
The next presenter was Gergely Timar, from Yahoo. His presentation, 10 lessons learned in the Trenches was not so detailed and practical, but there were good, general guidelines and specific cases were also presented which needed irregular solutions.
After this, we had a short break to have some drinks, have a coffee, and prepare for the next series of presentations. The third presenter, on day two, was Paul-Henning Kamp. His presentation, HTTP performance is a solved problem, was not a technical one, it was a more philosophical presentation with a lot of great questions asked related to users’ privacy, the need of privacy, who should be able to track and access peoples’ data on the web or on their mobile devices and what improvements should be made in this area.
Raymond Chen was the next presenter, I guess many people who have came from a windows background are familiar with his name. His presentation was kind of “out of the box” if we compare it to other ones. He approached the scalability concept from the community and developer point of view, which in their (Microsoft’s) case was much more a scale-down solution, simplify things, cleaning up APIs, redesigning interfaces..etc, when they were developing WinRT. His conception was interesting and he did present some interesting things, but I felt this more like a sponsor pitch than a presentation about scalability.
After Raymond’s presentation there was the lunch break, which passed by as nothing and we were founding ourselves listening to the next presenter, Andreas Ehn from Wrapp. In his presentation, Andreas focused on presenting the scaling issue from a founder, COO, CTO perspective. He presented a lot of ideas what other factors are affecting the scalability of an organization besides the technical ones.
The last presenter was a hungarian techie, Zoltan Toth-Czifra, working at Softonic in Barcelona, his presentation, Which is easier? Scaling from 100k to 10M or 10M to 1Bstarted a little boring, but by the end of his presentation it became very concrete and the advices which he gave were really useful to improve the scalability inside any company.
After the last speech there was a break and afterwards there were open sessions, one interesting topic was data sharding to increase scalability of systems, another topic, even more interesting, was Raspberry Pi from scalability point of view.
As a summary, all I can say, my overall impressions of the conference was very positive and if somebody would ask me, would I attend again, the answer without any hasitation would be, OF COURSE!!