Time |
Nickname |
Message |
18:02
🔗
|
soultcer |
Good idea, maybe even do both |
19:30
🔗
|
chronomex |
nah, actually there's little likelihood of much benefit |
19:30
🔗
|
chronomex |
clustering is only effective if you return many results at a time, so you cluster to reduce disk seeks while constructing large result sets |
19:31
🔗
|
soultcer |
Oh |
19:31
🔗
|
soultcer |
I didn't research how postgres actually stores the data. I know about toast and how the indexes work, but not the actualy data stuff or any of the MVCC stuff |
19:33
🔗
|
chronomex |
yeah |
19:34
🔗
|
chronomex |
I'm looking into clustering my db, since I serve out 30-40 results at once on many queries |
19:34
🔗
|
soultcer |
On what index are you clustering them? Country? City? Last name? |
19:35
🔗
|
chronomex |
city and name, I think. |
19:35
🔗
|
chronomex |
I have to get my city refs sorted out first though |