5 ms·

CMU 15-721: In-Memory Databases / Advanced Database Systems [video]

227 points by skovorodkin 7y ago

dowakin 7y ago
Also check out other courses from Prof. Andy Pavlo https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/playlists https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/pla...
calpaterson 7y ago
I find it hard to reconcile the incredible generousity of making this material available for free on the internet with the cringeworthiness of Andy Pavlo's style. I love the material but the "6th form humour" is really off-putting and he doesn't need it. See https://www.youtube.com/watch?v=m72mt4VN9ik&t=781 https://www.youtube.com/watch?v=m72mt4VN9ik&t=781 for an example of what I mean
- har777 7y ago
  I disagree. I thought it was hilarious :P
- thelastbender12 7y ago
  Ha! It really comes down to personal preferences. Distinctive style of presenting material is what really makes a lecture engaging to me. Also, as a selection filter given the abundance of content online.
- james_s_tayler 7y ago
  I love his style. He makes it fun.
- ahmedalsudani 7y ago
  Personally, I enjoy the material on databases—but if I’m being honest, the main reason I watch the first lecture is to find out what sort of trouble Andy Pavlo has gotten into.
- balfirevic 7y ago
  > he doesn't need it. Nobody needs anything. You didn't need to write your comment. And neither did I have to write this one. That's terrible criteria to judge anything, if it even means anything sensible at all.
  - rgoulter 7y ago
    I'm fine with "it's down to taste". "This course video doesn't need juvenile jokes" is an expression of taste. And some people enjoy the jokes.
  - rumanator 7y ago
    I have to agree with the parent poster: the skits do take away from the content. I'm grateful the author was so generous to share his knowledge with the world, but the skits are distracting and off-putting.
- ripley12 7y ago
  Andy’s gags are a tiny, tiny fraction of the material. Like, ~5min out of 10+ hours of lectures in a given class. They’re really easy to skip over if you don’t like them.
arkj 7y ago
That prof got sprayed on by a lady asking for 200 dollars (lecture 1, 13:00). Anybody knows the background?
- thepete2 7y ago
  That has to be a joke, maybe meant to keep your attention. He has made cuts so why else would he have kept it in?
- dominotw 7y ago
  i think its this from the top comment https://www.youtube.com/watch?v=m72mt4VN9ik&t=781 https://www.youtube.com/watch?v=m72mt4VN9ik&t=781
cnbscience 7y ago
Amazing material as always! feeling proud to be an alumni.
lichtenberger 7y ago
Great lecture so far. Haven't had the time to watch the whole lecture, but one thing I want to mention is that there are techniques to improve the buffer manager performance as for instance described here by Goetz Graefe: http://www.vldb.org/pvldb/vol8/p37-graefe.pdf http://www.vldb.org/pvldb/vol8/p37-graefe.pdf I've implemented an even simpler solution for my Open Source Data Store (https://sirix.io https://sirix.io) in that each page stores a number of references, which are itself lightweight pointer objects (in Java) and it simply stores an in-memory reference as well as a pointer to the location to fetch it from disk/a flash drive. If the buffer manager has these Objects as keys on eviction we can simply null the reference to the in-memory page instance.
manigandham 7y ago
Check out the CMU database group for all the other content and the multiple courses: https://db.cs.cmu.edu/ https://db.cs.cmu.edu/ Here's their Youtube channel: https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/featured https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/fea...
jules 7y ago
I've watched some of these and the material and teacher are awesome. Two questions come to mind: 1. A ton of effort seems to be spent on making things run in parallel, but that introduces quite a bit of overhead too, so how well does a sequential baseline actually perform? By sequential baseline I mean a single thread that just executes all incoming transactions one by one in sequence. 2. This course seems to spend a lot of time on things that the teacher says are things you shouldn't do anyway. For instance there is an entire lecture on skip lists and Bw-trees, and at the end the teacher mentions that these are terrible. This is interesting from a historical perspective, but not only does this take a lot of time, I also lose track of which things you should and which things you shouldn't do. It'd be interesting to have a compressed course that spends less time on things you should not do, perhaps by adding annotations to the video to skip sections that are about things you should not do.
- twoodfin 7y ago
  A ton of effort seems to be spent on making things run in parallel, but that introduces quite a bit of overhead too, so how well does a sequential baseline actually perform? By sequential baseline I mean a single thread that just executes all incoming transactions one by one in sequence. You should check out the H-Store research project[1] and its commercial successor VoltDB. They’re basically a study in how much you can win with a federation of single-threaded database systems. [1] https://hstore.cs.brown.edu/papers/hstore-endofera.pdf https://hstore.cs.brown.edu/papers/hstore-endofera.pdf
- rohansuri 7y ago
  Do you have the link to lecture on SkipLists and Bw Trees? Thanks
  - jules 7y ago
    Lecture 7.

227 points by skovorodkin 7y ago

Also check out other courses from Prof. Andy Pavlo https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/playlists https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/pla...

calpaterson 7y ago

I find it hard to reconcile the incredible generousity of making this material available for free on the internet with the cringeworthiness of Andy Pavlo's style. I love the material but the "6th form humour" is really off-putting and he doesn't need it. See https://www.youtube.com/watch?v=m72mt4VN9ik&t=781 https://www.youtube.com/watch?v=m72mt4VN9ik&t=781 for an example of what I mean

har777 7y ago
I disagree. I thought it was hilarious :P
thelastbender12 7y ago
Ha! It really comes down to personal preferences. Distinctive style of presenting material is what really makes a lecture engaging to me. Also, as a selection filter given the abundance of content online.
james_s_tayler 7y ago
I love his style. He makes it fun.
ahmedalsudani 7y ago
Personally, I enjoy the material on databases—but if I’m being honest, the main reason I watch the first lecture is to find out what sort of trouble Andy Pavlo has gotten into.
balfirevic 7y ago
> he doesn't need it. Nobody needs anything. You didn't need to write your comment. And neither did I have to write this one. That's terrible criteria to judge anything, if it even means anything sensible at all.
- rgoulter 7y ago
  I'm fine with "it's down to taste". "This course video doesn't need juvenile jokes" is an expression of taste. And some people enjoy the jokes.
- rumanator 7y ago
  I have to agree with the parent poster: the skits do take away from the content. I'm grateful the author was so generous to share his knowledge with the world, but the skits are distracting and off-putting.
ripley12 7y ago
Andy’s gags are a tiny, tiny fraction of the material. Like, ~5min out of 10+ hours of lectures in a given class. They’re really easy to skip over if you don’t like them.

arkj 7y ago

That prof got sprayed on by a lady asking for 200 dollars (lecture 1, 13:00). Anybody knows the background?

thepete2 7y ago
That has to be a joke, maybe meant to keep your attention. He has made cuts so why else would he have kept it in?
dominotw 7y ago
i think its this from the top comment https://www.youtube.com/watch?v=m72mt4VN9ik&t=781 https://www.youtube.com/watch?v=m72mt4VN9ik&t=781

cnbscience 7y ago

Amazing material as always! feeling proud to be an alumni.

lichtenberger 7y ago

Great lecture so far. Haven't had the time to watch the whole lecture, but one thing I want to mention is that there are techniques to improve the buffer manager performance as for instance described here by Goetz Graefe: http://www.vldb.org/pvldb/vol8/p37-graefe.pdf http://www.vldb.org/pvldb/vol8/p37-graefe.pdf I've implemented an even simpler solution for my Open Source Data Store (https://sirix.io https://sirix.io) in that each page stores a number of references, which are itself lightweight pointer objects (in Java) and it simply stores an in-memory reference as well as a pointer to the location to fetch it from disk/a flash drive. If the buffer manager has these Objects as keys on eviction we can simply null the reference to the in-memory page instance.

manigandham 7y ago

Check out the CMU database group for all the other content and the multiple courses: https://db.cs.cmu.edu/ https://db.cs.cmu.edu/ Here's their Youtube channel: https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/featured https://www.youtube.com/channel/UCHnBsf2rH-K7pn09rb3qvkA/fea...

jules 7y ago

I've watched some of these and the material and teacher are awesome. Two questions come to mind: 1. A ton of effort seems to be spent on making things run in parallel, but that introduces quite a bit of overhead too, so how well does a sequential baseline actually perform? By sequential baseline I mean a single thread that just executes all incoming transactions one by one in sequence. 2. This course seems to spend a lot of time on things that the teacher says are things you shouldn't do anyway. For instance there is an entire lecture on skip lists and Bw-trees, and at the end the teacher mentions that these are terrible. This is interesting from a historical perspective, but not only does this take a lot of time, I also lose track of which things you should and which things you shouldn't do. It'd be interesting to have a compressed course that spends less time on things you should not do, perhaps by adding annotations to the video to skip sections that are about things you should not do.

twoodfin 7y ago
A ton of effort seems to be spent on making things run in parallel, but that introduces quite a bit of overhead too, so how well does a sequential baseline actually perform? By sequential baseline I mean a single thread that just executes all incoming transactions one by one in sequence. You should check out the H-Store research project[1] and its commercial successor VoltDB. They’re basically a study in how much you can win with a federation of single-threaded database systems. [1] https://hstore.cs.brown.edu/papers/hstore-endofera.pdf https://hstore.cs.brown.edu/papers/hstore-endofera.pdf
rohansuri 7y ago
Do you have the link to lecture on SkipLists and Bw Trees? Thanks
- jules 7y ago
  Lecture 7.