CODEWITHSUNDEEP
cassandra
Ofek Ron
Ofek Ron

Reputation: 8580

Cassandra wide row and simple row behind the scenes

Trying to learn how cassandra works here, was very hard to get insights about how data is stored behind the scenes, hence im not sure i got it right and wanted to get corrected if i got it wrong, so far i understand that a typical row that the partition key (leftmost column or compuond column in the primary key) is unique, a row would be written to disk in a chunk that looks like this :

enter image description here

where the columns are sorted by their names.

But, if the partition key is not unique for then it would be considered a "wide row" and a row would look like the following examples :

enter image description here

Please correct me if i got it wrong...

Upvotes: 2

Views: 79

Answers (1)

undefined_variable
undefined_variable

Reputation: 6218

For Second part where you have compound primary key the structure will be somthing similar to first except for the fact that:

ColumnName will be replaced by usrname1.comments|username1.comments_ts|username2.comments|username2.comments_ts

VideoId : usrname1.comments|username1.comments_ts|username2.comments|username2.comments_ts

Same thing will be true for comments_by_user

username: videoid1.comments|videoid1.comments_ts|videoid2.comments|videoid2.comments_ts

PS: I am not good at drawing images so you have to do with this text answer.

For more details Refer Slide 48

Upvotes: 2

Related Questions