I am really excited to present Crystal at the VLDB 2021 live in Copenhagen (Session 44). We also recorded a video for those who cannot attend [1]. Badrish Chandramouli Our SIGMOD/PODS 2021 talk on JSON Tiles is also available online [2]. [1] youtu.be/bI0_1SVn1ZM [2] youtu.be/SgkupIfUSWY

1/2 Caching for analytics has resurfaced as a challenge in today's disaggregated cloud analytics database architecture. At @VLDB2021, we present research on Crystal - a new DB-agnostic semantic caching middleware. Paper: badrish.net/papers/crystal… Video: youtube.com/watch?v=bI0_1S…

2/2 Crystal uses local storage at compute nodes to accelerate queries and increase the value per cached byte. We show Crystal accelerating queries over Parquet, on unmodified Spark & Greenplum, with small data sources. Joint work w Microsoft Research intern Dominik Durner & Yinan Li.

📙 Crystal: A Unified Cache Storage System for Analytical Databases Authors: Dominik Durner, Badrish Chandramouli, Yinan Li Featured in PTK #92 ​​bit.ly/ptk92 Paper: vldb.org/pvldb/vol14/p2…

CedarDB: The 🇩🇪German-powered, PostgreSQL-compatible freak-of-nature database management system based on TUM's Umbra (Thomas Neumann + team) is out of stealth and now available: cedardb.com/blog/ode_to_po… /cc CedarDB

Super happy to share what we've been up to the last months. Today, we launched our blog.

Database systems need simple, efficient, and robust hash tables for join processing! We spill all our secrets in a paper & in the first tech deep dive on our blog: cedardb.com/blog/simple_ef…

Our new blog post gives detailed insights into the inner workings of our hash tables. Enjoy the read!

Your SSD (probably) sucks! Find out why and how your database system can deal with it in our latest tech deep dive at cedardb.com/blog/ssd_laten…

Have you ever wondered how you can use a database system for good? Then dive into our latest blog post and learn how to work with JSON data and graphs in CedarDB while hunting for Germany's most wanted white-collar criminal! cedardb.com/blog/working_w…

We were surprised to find out that everybody uses our strings. They even call them "German Strings". 😲🇩🇪 Take a look at our latest blog post to find out why DuckDB, ApacheArrow, polars data, and Facebook Velox use our custom string implementation. cedardb.com/blog/german_st…

Vol:16 No:11 → Exploiting Cloud Object Storage for High-Performance Analytics vldb.org/pvldb/vol16/p2…

I am happy to share our blog post about our new hybrid storage engine that allows us to process analytical and transactional workloads. Details of our storage engine will also be presented at VLDB'24!

For a high-level overview of how we achieve great OLAP and OLTP performance, you can also browse our blog post at cedardb.com/blog/colibri/. If you are at VLDB this year, don't miss Tobias' presentation ;)

Query optimization is one of the hardest tasks in database systems. Get a quick overview of how our join ordering algorithm is able to unnest arbitrary queries.

Why old-school exceptions are still a good idea: They provide a better user and developer experience and are even faster than error values! Read more in our blog: cedardb.com/blog/exception…

Having data in memory is considered the gold standard for data analysis, but are you sure that RAM locality alone is enough? Read on to learn about the new king of efficient data processing: cedardb.com/blog/reducing_…

The underlying data layout of your program can either help or hurt your algorithms. In our latest post, we explore why optimizing the storage layout is the key to unlocking blazingly fast data processing. cedardb.com/blog/optimizin…

Happy to announce that we released our community edition! Built for speed, CedarDB delivers top-tier performance for every workload, from ultra-responsive OLTP to large-scale OLAP - and everything in between.

Leaving academia is always a big step, especially if you bring your research project with you into the real world. Read our latest post to learn what we did to prepare a research project for production workloads and what we learned along the way: cedardb.com/blog/research_…