Jan K.

[Carsten Binnig @cbinnig]

We just opened up the submission site of @ACMSoCC 2022. So, consider submitting your work. We also have a new work-in-progress track to discuss early-stage results of highly promising research. The submission deadline is June 17. Abstract registration is June 10.

[Hyrise @hyrise_db]

Interested in how Reinforcement Learning can be used for workload-aware budget-constrained index selection? We will present our SWIRL paper tomorrow at 11am and 9pm (MEST) at @edbticdt2022. Get the paper now at https://t.co/WUBP1pJqZ4

[Jan K. @Jan_Kossmann]

After two years of virtual talks, I am glad that I got the chance to give another live presentation towards the end of my PhD journey at @cidrdb today. Check out our paper about Data Dependencies for Query Optimization & our research vehicle @hyrise_db - https://t.co/8mVsJzzlul

[Hyrise @hyrise_db]

Using data dependencies for query optimization is challenging in practice. Our @cidrdb paper "Workload-driven, Lazy Discovery of Data Dependencies for Query Optimization" in collaboration with @HPI_DE's "Information Systems Group" (https://t.co/D395Ltepb0) is here to help.

[Hyrise @hyrise_db]

Thank you Benjamin Wagner for yesterday’s insightful vaccination database talk (@andy_pavlo & @CMUDB) on how @FireboltHQ is built. In particular, we are very proud to hear that Firebolt’s query optimizer and parser are forked off Hyrise.

[EPIC Research Group @HPI_EPIC]

After building minimal databases, the students of our 'Develop your own Database' course present their results tackling issues such as Cardinality Estimation & Partial Indexes in our open-source @hyrise_db.
@HPI_DE @Jan_Kossmann @martinboissier
https://t.co/V8aXYG28Uh

[Hyrise @hyrise_db]

Using data dependencies for query optimization leads to more efficient database query plans. We published a survey of these techniques in collaboration with @HPI_DE's "Information Systems Group" (https://t.co/D395Ltepb0) in the VLDB Journal: https://t.co/prwy3a9dL0 @Jan_Kossmann

[Hyrise @hyrise_db]

We compared Apple’s M1 chip (MacMini) with recent AMD and Intel CPUs. We evaluated TPC-H with scale factor 1.0 in Hyrise. Both single- and multi-threaded performance are impressive.

1/2

[Hyrise @hyrise_db]

Magic mirror in my hand, which is the best in the land? The index selection problem has been researched for the past 50 years. Our “Experimental Evaluation of Index Selection Algorithms” sheds light on above’s question by evaluating 8 different approaches. https://t.co/OPIka3gwHf

[PVLDB @pvldb]

Vol:13 No:11 → Magic mirror in my hand, which is the best in the land? An Experimental Evaluation of Index Selection Algorithms https://t.co/1UfPvalXCu

[Jan K. @Jan_Kossmann]

Seltsam, dass diese haltbare Gläschennahrung aktuell so beliebt zu sein scheint 🧐 Bitte lasst für alle etwas übrig! https://t.co/7ajoQffasZ

[Buttercup Kittensnatch @StereoSushisu]

Langsam gehen Artikel rum, wie man sich am besten auf eine Pandemie vorbereitet. Dort werden Hamsterkäufe empfohlen. Bitte bleibt ruhig und denkt daran, dass jeder einzelne Hamster in artgerechter Haltung ganz schön viel Grundfläche braucht. Vielleicht alle erstmal nur einen.

[Hyrise @hyrise_db]

Recently, we merged the new server code that Toni wrote for his master's thesis. It improves network request handling (buffer and thread management, efficient serialization) and connection management. Almost an improvement of 600x no-op throughput.

[Christoph Matthies | @[email protected] @chrisma0]

In the newest blog post of our @HPI_DE master's project on @hyrise_db the students detail how @Intel's performance profiler VTune Amplifier was used to find parallelization optimizations, resulting in an average performance boost of around +40 percent across TPC-DS queries.

[Christoph Matthies | @[email protected] @chrisma0]

Find out how the "Disjunction Split-Up" optimizer rule of @hyrise_db improved a TPC-DS database query runtime by 22,000%. 📈🖥️🔍
The most recent @HPI_DE master's project blog post:
https://t.co/gQvIZyL5vU

[Hyrise @hyrise_db]

Our master's project published another blog post (two weeks ago, sorry)! This time about Sub-Plan Memoization. The post also includes some interesting comparisons against @DuckDB2. Check it out!

https://t.co/GgnyEqBM9r

[Jan K. @Jan_Kossmann]

@andy_pavlo @martinboissier @hyrise_db @CMUDB What do you think is the problem with our plugin approach? I would be curious for your thoughts on that. Maybe we could have a chat at the AIDB workshop at this year's VLDB?

[Hyrise @hyrise_db]

Our current master’s project works on bringing the TPC-DS benchmark into Hyrise. The newest blog post contains a detailed summary of the benchmark’s specification, its design decisions, and history: https://t.co/dKmR78Uptj

[Ann Rosenberg @rosenbergann]

Thanks @HPI_DE for teaming up on the #GlobalGoals @SAPTechEd @SAPMentors

[Christoph Matthies | @[email protected] @chrisma0]

What do Scooter references have to do with Database statistics? Find out in the latest #hyrise master’s project post
https://t.co/sAgcrFt4Ic

[Christoph Matthies | @[email protected] @chrisma0]

The 3rd post of our current master’s project on the "brain of every database", the SQL optimizer, is out. #hyrise
https://t.co/zvltHByssU