Before Summingbird at Twitter, users that wanted to write production streaming aggregations would typically write their logic using a Hadoop DSL like Pig or Scalding. These tools offered nice distributed system abstractions: Pig resembled familiar SQL, while Scalding, like Summingbird, mimics the Scala collections API. By running these jobs on some regular schedule (typically hourly or daily), users could build time series dashboards with very reliable error bounds at the unfortunate cost of high latency. While using Hadoop for these types of loads is effective, Twitter is about real-time and we needed a general system to deliver data in seconds, not hours. Twitter’s release of Storm made it easy to process data with very low latencies by sacrificing Hadoop’s fault tolerant guarantees. However, we soon realized that running a fully real-time system on Storm was quite difficult for two main reasons: Recomputation over months of historical logs must be coordinated with Hadoop or streamed through Storm with a custom log loading mechanism; Storm is focused on message passing and random-write databases are harder to maintain. The types of aggregations one can perform in Storm are very similar to what’s possible in Hadoop, but the system issues are very different. Summingbird began as an investigation into a hybrid system that could run a streaming aggregation in both Hadoop and Storm, as well as merge automatically without special consideration of the job author. The hybrid model allows most data to be processed by Hadoop and served out of a read-only store. Only data that Hadoop hasn’t yet been able to process (data that falls within the latency window) would be served out of a datastore populated in real-time by Storm. But the error of the real-time layer is bounded, as Hadoop will eventually get around to processing the same data and will smooth out any error introduced. This hybrid model is appealing because you get well understood, transactional behavior from Hadoop, and up to the second additions from Storm. Despite the appeal, the hybrid approach has the following practical problems: Two sets of aggregation logic have to be kept in sync in two different systems; Keys and values must be serialized consistently between each system and the client. The client is responsible for reading from both datastores, performing a final aggregation and serving the combined results Summingbird was developed to provide a general solution to these problems.Very interesting stuff. I’m particularly interested in the design constraints they’ve chosen to impose to achieve this — data formats which require associative merging in particular.
We mentioned above that pop-up spaces have become popular across Europe because they allow developers and city councils to harness urban creativity in order to drive up real estate prices without ceding control of a given site. Those who produce the space through hard work, collaboration and passion move on, making way for property development and speculation. The international research in this area is very clear on this point and it has been documented in places from Lower-East Side Manhattan to Berlin’s Kreuzberg. Most perversely, increased property prices make it even more difficult for creativity to flourish in a given area and end up driving out long-term working class communities, migrants and young people. But what can we do? If every attempt we make to make our city a better place simply ends up being captured in the calculations of real estate players, surely the situation is hopeless? Is it better, then, to do nothing? We don’t think it is better to do nothing and, like Upstart, we still believe we can find a way together through experimentation and collaboration. However, this means questioning, reflecting on and publicly discussing the relationship between our efforts to make a city more after our hearts desire and the process of gentrification. As noted above, this is especially the case with pop-up spaces given their temporary nature. It is really necessary that we think about how to make sure our activities don’t contribute to gentrification in the long term, but instead benefit the city as a whole. We certainly don’t have the solutions, but if we sweep these awkward questions under the carpet we risk contributing to the very forces we want to challenge and alienating those who will perceive us as the ‘front-line’ of gentrification.
Disruptor: decimating P99s since 2011
I love these.
Photographic prints are great because they don’t need power to be displayed. They are more or less permanent. Videos are great because they record a sequence of time which shows reality almost like how we experience. Is it possible to combine the two? And not via long exposure photography where often details are lost from motion. So I played around with the tools of digital photography and post processing to give you this series: Time is a dimension. This series of images are mostly landscapes, seascapes and cityscapes, and they are a single composite made from sequences that span 2-4 hours, mostly of sunrises and sunsets. The basic structure of a landscape is present in every piece. But each panel or concentric layer shows a different slice of time, which is related to the adjacent panel/layer. The transition from daytime to night is gradual and noticeable in every piece, but would not be something you expect to see in a still image.
Links for 2013-09-03
permalink. Both comments and trackbacks are currently closed.. Bookmark the