It is really demystified. Easy to understand. Need more videos like this. Wish you'd have had more feedback and appreciation from the audience to continue making these videos.
Hi Michael, thank you for sharing your knowledge. Rather than solely providing theoretical explanations, it would have been helpful to delve into practical examples, such as querying different columns for a record in columnar databases. Given that data (columns) in columnar databases is scattered across various blocks, it's crucial to discuss the unique identifier in such scenarios.
Very Good explaination. I have one question, How do you handle the row insert operation in case of Column oriented database as new data will need to inserted into each block.
Good question! This is non-trivial and why many data warehouses either don't allow or strongly discourage row-level inserts and updates (since you have to figure out where in each block to insert everything and move other things around).
Good explanation but in practice it's very rare to query single columns so the benefits of columnstore are much smaller than expected. I work in a table with 100 million rows, moved from rowstore to columnstore and didn't notice that great improvement; on the contrary after enabling columnstore I still had to create quite a few non clustered indexes to cover my queries
Require your help in understanding In case Where there is a table with multiple records of a date And we write the below query : Example select name, id, date From table Where date =2017 In which row / columnar it will perform better If there was just one record of 2017 then it would definatly be row store But in case where there are more tha 1 row of the record which platform works better Note - there are multiple records of 2017
Here row store index will be highly useful and generally u will be using Id as clustered index so now u can go with date as non clustered index so now all dates are arranged now u can access all rows with 2017 much easily. But when u take column store index data will be separated as columns so it would be take time to access all columns for a row by checking many blocks.
The concepts are clearly explained! Love the way you structure the content.
Thank you!!
This is the best breakdown I've ever seen of this topic. Two thumbs up!
This is the best video I’ve found to explain this. Thank you!
Nice job. Excited for what's coming here!
Great explanation of this concept, thank you!
Should have much more views and likes!! best explanation ever
Thank you so much! Definitely check out my other videos as well if you want to go deeper on other important database topics :)
Hands down the best video on this topic.
You were amazing with your crystal clear explanation! Kudos!
It is really demystified. Easy to understand. Need more videos like this. Wish you'd have had more feedback and appreciation from the audience to continue making these videos.
Great explanation tying the db internals to the benefits and drawbacks of row and columnar databases.
Thank You :) This is BEST crash course ever!
I am your 1000th Subscriber. Keep up the good work.
One of the best explanations I've come across. Thanks !
simple yet detailed explanation. thank you
Jeez, what a great explanation. More so, in as little as 6 minutes!
Amazing presentation Thanks!
Very nice explanation, thanks
Excellent! Great explanation that's really helped me
Thank you for such classic breakdown of this concept
Hi Michael, thank you for sharing your knowledge. Rather than solely providing theoretical explanations, it would have been helpful to delve into practical examples, such as querying different columns for a record in columnar databases. Given that data (columns) in columnar databases is scattered across various blocks, it's crucial to discuss the unique identifier in such scenarios.
Real good explanation!!
very simple and clear way of explanation of ROW Vs Column Store index - please make more videos-
One of the best explanations. Thanks 👍
Great explanation, thank you so much
Holy shit, what a great video. Been looking for this explanation for a long time
Pretty briefed explanation. Need reference for Big Data, Data warehouse, etc. Thanks, man!
Very nice video. Clearly explained, short and efficient! Loved it (y)
excellent explanation.
Thank you, that clearly explained what is the column store
this is just amazing in such a short video how clearly you explained everything
This video is Gold!
Excellent explanantion...!!
best video ever on the topic
Very clear and excellent! Thanks!
So easy to understand. Thanks
Perfect video! Brilliant explanation!
Great video
Very Good explaination. I have one question, How do you handle the row insert operation in case of Column oriented database as new data will need to inserted into each block.
Good question! This is non-trivial and why many data warehouses either don't allow or strongly discourage row-level inserts and updates (since you have to figure out where in each block to insert everything and move other things around).
Basically the answer is that "it's a PITA" for all of the reasons you can imagine.
Thank you sir .. Got the Answer. spot on...
Good explanation but in practice it's very rare to query single columns so the benefits of columnstore are much smaller than expected. I work in a table with 100 million rows, moved from rowstore to columnstore and didn't notice that great improvement; on the contrary after enabling columnstore I still had to create quite a few non clustered indexes to cover my queries
What sorts of queries were you comparing between the two systems? And yes, indexes definitely still matter!
Great explanation! Thanks for this video!
Very good !!
Thanks for this, explained well
Thanks for this!
wow didnt know usyk was into data.. great video too!
Great explanation, thank you!
Well explained
Brilliant !!
Will keep watching these vids, but I'm wondering if anyone has any book suggestions that has all of these concepts explained well and comprehensively.
greate video
Store data by rows or.
Store data by columns.
Column storage is good for analytical workloads.
Column storage is easy to compress.
Nice
🤯🤯🤯
Require your help in understanding
In case
Where there is a table with multiple records of a date
And we write the below query :
Example
select name, id, date
From table
Where date =2017
In which row / columnar it will perform better
If there was just one record of 2017 then it would definatly be row store
But in case where there are more tha 1 row of the record which platform works better
Note - there are multiple records of 2017
Here row store index will be highly useful and generally u will be using Id as clustered index so now u can go with date as non clustered index so now all dates are arranged now u can access all rows with 2017 much easily.
But when u take column store index data will be separated as columns so it would be take time to access all columns for a row by checking many blocks.
Great explanation, thank you.
Thank you!