Consider the following HBase table design for blogs. On the website, logged in users (who all have a unique, integer…
Consider the following HBase table design for blogs. On the website, logged in users (who all have a unique, integer user ID) may comment on blog posts.
| cf: meta | cf: comments | |||
| post11356 | user: 3 | content: … | 1: <uid, comment> | 2: <uid, comment> |
| post11357 | user: 7 | content: … | ||
| post11358 | user: 5 | content: … | 1: <uid, comment> | |
| post11359 | user: 3 | content: … | 1: <uid, comment> | 2: <uid, comment> |
For each query below, state either YES or NO the schema is optimal for that query. If you state NO, re-design the schema (copy the table above and modify it, then explain in English what is different with the new schema) so it is optimal for that specific query.
- a) Given a particular blog post, what are all of the comments associated with it?yes
- b) What are the blog posts for a particular user?
- c) What comments has a particular user posted?
- d) Deleting a particular blog post and all related data.
