You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a plan with a simple Seq Scan that seems innocent, but takes too much time (see Mantis 10932).
The reason is this:
Loops: 170252550
It is not obvious enough.
Usually the problem is a Nested Loop, that is horribly inefficient if done too many times.
The node is tagged Slow and shows a high buffers count, but the Nested Loop may be out of the screen and a user may wonder why it is so for a single node if it overlooks the Loops line.
But I'm not sure which rule to use.
An arbitrary rule would be better than nothing:
more than 5 loops -> yellow ;
more than 100 -> red ;
more than 1000 -> brown
The text was updated successfully, but these errors were encountered:
I have a plan with a simple Seq Scan that seems innocent, but takes too much time (see Mantis 10932).
The reason is this:
It is not obvious enough.
Usually the problem is a Nested Loop, that is horribly inefficient if done too many times.
The node is tagged Slow and shows a high buffers count, but the Nested Loop may be out of the screen and a user may wonder why it is so for a single node if it overlooks the Loops line.
But I'm not sure which rule to use.
An arbitrary rule would be better than nothing:
more than 5 loops -> yellow ;
more than 100 -> red ;
more than 1000 -> brown
The text was updated successfully, but these errors were encountered: