Відео

Either that blockchain guy is a troll or that was one of the most entitled douches I've ever seen at a lecture.

superb bro. Loved ur lecture

For the log structure, if we still need an index for look up, how to save the index? How updating that index does not end up in random io stuff?

really well explained! thank you!

what is the outro song??

Why is the "lost updates" anomaly missing in the discussion of isolation levels?

this is f*cking awesome

Great lecture.. but for this specific lecture, camera is moving too much and also quality is not HD.

Thank you for the lecture. Can you please speak a little slow next time or add subtitles? Microphone doesn't do justice in picking up your enunciation of certain words.

insightful lecture

In 1:20:32, Oracle/MySQL and Postgres don't use memory as the primary storage, do they? And with that, Oracle/MySQL still beat most in-meomry DBMS? Is it because their WAL was disabled for this benchmark?

I have a question with regard to the AllocatePage method in bustub when doing projec 1(can't paste link because it keep getting deleted). The way it try to retrieve a new page id is just return next_page_id_++ with next_page_id_ initialized to 0? But should it not consider what other page ids already used in disk? I look around in the codebase but did not find any code try to keep track of the page ids used in disk. Please let me know if I miss anything, thanks a lot

This is how you do it in Oracle btw SELECT TO_DATE ('2023-08-30', 'YYYY-MM-DD') - TO_DATE ('2023-01-01', 'YYYY-MM-DD') AS days FROM DUAL;

thanks a lot!

isn't sql would do a linear scan when searching for a tuple ?

The lectures are undoubtedly brilliant. But, I really really love the intro and the outro too. What a production! So cool.

What about the pure MOLAP’s and financial modeling databases like Essbase, Hyperion, TM1. Plus Cognos PowerPlay, SQL Server Analysis Services?

Just clarification on Sybase IQ - this was/is a fully fledged columnar store, not just an in memory accelerator. Sybase did build a product like that, called RAP (Realtime Analytics Platform) where they used their row store in memory, and IQ as a backing column store for analytics.

thanks a lot.

8:41

kraska is not a good guy

Is this course any different from the one from 2018?

just started the course , and I can tell , it's gonna be thick haha

Missing is an understanding of the optimizer and how to debug things. When things go wrong, when predicates are not pushed down. You ask why? And you have nothing.

audio really is bothering me but i have to learn

thanks a lot!

I'm surprised that Netezza didn't come in for a mention for the mid-2000 based Postgres based OLAP systems, it's very relevant because the AQUA FPGA concept was already in user by the Netezza appliances back the mid-2000s. AQUA had super poor adoption and they never announced that it went away.

thanks

nice explanation of Redhsift. I could not understand Primitive supported by Redshift only not by other databases. when I search i do see that Snowflake support all 8 primitive but not supported array something. if you can help to explain more on this please. thanks

DPDK sounds like an in-house version of QUIC for reliable data transmission over UDP en.wikipedia.org/wiki/QUIC

Nice to see you have moved bloom filters from join algorithms section last year to general hash tables lecture, cool staff

thanks a lot. really appreciated.

needs more details on how group by is done in volcano/iterator model

@31:08 correction: dataframes came from S, then R, and then pandas. Wes talks about how he stole them for Ibis/Arrow/pandas in another video I'm forgetting. Sincerely, - R user

wow the blockchain idiot!! 😮 I really can't believe there are people like these in CMU😵‍💫

thanks a lot.

Another little detail: spark can delegate saving of shuffle data to the Hadoop Yarn NodeManager process -which can serve data even after the spark worker process terminates. This allows for more agile spark clusters within a Hadoop cluster. However, with the move to kubernetes container hosting spark serves the data itself and assumes that it won’t terminate. This is potentially a problem with deployment on spot-priced cloud VMs as the “your server is fairly reliable” no longer holds.…

Excellent explanations/lecture Prof Andy!

Just don't use NULL, there can't be a data type for an absence of data (even more contradictory the absurd of a phrase like "null value") . The relational data model itself does not have NULL, it is based on 2 valued logic, not on 3 valued logic. I do think this is an awesome course but in general it does not properly observe the theoretical background for the relational data model.

Thanks dr.Andy,I wish for you eternal happiness

To my knowledge IDS has been used in industry for a long time. See Bull's GCOS with IDS/II.

GREAT PRESENTATION! VERY CLEAR! 🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠🧠

I don’t work with Snowflake in my current role but I think about the white paper frequently. “Cloud native” is thrown around so loosely but it wasn’t until that white paper where I had seen something that was truly cloud native from start to finish and was only possibly with the scale and flexibility of modern cloud providers.

Impressive the throughput of content Justin inserts into his talk! - very smooth, quite impressive! - Happy to have learn about BigQuery

Second, distributed, here to stay

does anyone know how they handle security in JIT systems like hyper/umbra? like how can they stop someone from getting the compiler to generate code that reads or writes the system's memory?

These lectures should've followed HDFS....2011...instead of 2024.....with Compute/Security/Cost/etc, "Cloud should gradually depreciate", doesn't need to join the "Taxing Community..."

What is the worst idea?

Here first, fastest. I must be vectorized.

Here first, fastest. I must be vectorized.