Wouter Zorgdrager retweeted
Took >3 years with @MarioFragkoulis, @SenorCarbone & @vkalavri to write a survey on the evolution of streaming systems in the last 20 years (@VLDBconf J.).
Looking for a single point of reference for all things streaming systems research? It’s here:
link.springer.com/article/10…
Wouter Zorgdrager retweeted
Time and again, I 've found that the only way to get performance out of a system is to avoid doing work (often at the expense of memory). Concurrency is only to be applied when 1. the problem is amenable, 2. all other optimisations have been explored in the non-parallel code path
Wouter Zorgdrager retweeted
Please retweet!
I am searching for two PhD students and one postdoc for my VIDI grant. We are building a programming model & self-driving dataflow engine for Transactional Stateful Functions.
For PhD apply here by 15th July: tudelft.nl/over-tu-delft/wer…
Postdoc: drop me an email!
Wouter Zorgdrager retweeted
At around 17:30 Amsterdam time at @cidrdb I will be presenting how one can take an object-oriented Python programs and compile it to a dataflow graph and deploy it on streaming dataflow engines. (cidrdb.org/cidr2023/papers/p…).
w @kPsarakis,@wouterdz,@MarioFragkoulis, and @guidosalva.
I'm thrilled to to announce that I was awarded the #VIDI grant from @NWONieuws! The funding will allow me to build a transactional actor-like Cloud programming model on top of dataflow systems!
Wouter Zorgdrager retweeted
If you are a fan of homomorphic encryption, then try your hand at the secure biometrics competition @daniel_a_vos and I are organizing! Registrations opened today and the data becomes available on Friday next week. (And yes, it involves dogs 🐕)
More at: cyscompetition.com
Wouter Zorgdrager retweeted
(1/2) On Thu at 10:45 CET (@icdeconf), Jim Verhijde (w/ @MarioFragkoulis, @bkarak) will present how internal stream-processor state can be queried efficiently, and a stream processor with queryable state can replace caching/view layers in scalable data processing architectures.
Wouter Zorgdrager retweeted
Please RT: I am searching for a postdoc (funding for 2-3 years) to work on compiling actor-like programs into transactional dataflow graphs and executing them in dataflow systems. General project motivation in the video below and this preprint:
arxiv.org/abs/2112.00710
I uploaded my talk at #flinkforward here:
youtu.be/jMmEDPGeh0w
TL;DW; Building scalable cloud apps is a pain in the butt. To scale out, people actually build parallel dataflow systems without knowing it. This needs to change. We are working on it.
Wouter Zorgdrager retweeted
🔬Over the past few months, I’ve been working on TestAxis: a tool that gives you insights into failing tests in CI builds.
🧑🔬 Are you curious to check out TestAxis and help us make it better? Join our experiments!
testaxis.io/experiment
Thread ⬇️
Wouter Zorgdrager retweeted
Dear software practitioners,
If you have been involved in the selection of software libraries in your projects, we would appreciate it if you could help us participating in this short survey: surveygizmo.com/s3/5327059/1…
Thanks a lot, and feel free to RT.
Wouter Zorgdrager retweeted
Happy to see @ghtorrent as a prominent member of the newly announced @github Archive program, along the likes of @SWHeritage and Github Archive! archiveprogram.github.com/
Wouter Zorgdrager retweeted
Fantastic news to those running @ApacheFlink in #Kubernetes: @lyft has generously released github.com/lyft/flinkk8soper…. Job lifecycle management no longer such a pain! Many thanks to @anand12100, @ketanumaren and team.
Wouter Zorgdrager retweeted
A Brief History of Flink: Tracing the Big Data Engine’s Open-source Development by Alibaba Tech link.medium.com/r6mdVZpksU
Wouter Zorgdrager retweeted
Snow today for @FastenProject second day of the kick-off meeting in @tudelft Science Center (beautiful room btw).
