It is tempting to think of it as some kind of container. But we need to be careful here: the containers with which we are familiar are physical objects, and on this view space is not a physical object.
It is tempting to think that space is the void. But how does this differ from there being no such thing as space? How can there be such a thing as the void? And how can something physical be in it? How could it have features of its own? (Space does have features: it is curved, for example, according to modern physics.)
Another view is that space is a plenum: it is full of energy. Modern physics even says that a vacuum is in a highly energetic state - no part of space is not suffused with energy. But the field equations of Einstein's general theory of relativity allow that space can be devoid of all matter and energy.
Question: What then determines the boundaries of space? Example of a world in which just three objects are arranged at the vertices of an equilateral triangle - where are the boundaries of space in this world?
Lowe's preferred version of relationalism: space consists of all the locations that it is possible for a material object to be located. What determines the extent of space is thus facts about possible movements of material objects. These are determined by the natural laws of the universe. He calls this a modal version of relationalism.
On one way of understanding it, it says that material objects are just deformations in the fabric of space. This suggests that absolutism is right. Einstein's field equations also allow that space is perfectly empty.
In the experiment there is a bucket which is half full of water, hanging by a rope. In stage 1 it is stationary, and there is no relative motion between the water and the bucket. In stage 2 the bucket is spun, and there is relative motion between the water and the bucket. In stage 3 the water catches up to the bucket, and there is again no relative motion between the water and the bucket.Newton says that in stage 2 the water is rotation with respect to the bucket, but it is not really rotating, whereas in stage 3 it is really rotating, and that is why its surface becomes concave. We can detect this real rotation (by looking at the surface), and thus detect real acceleration, and from that infer the existence of real velocity.
Perhaps we might try the following analogy: just as there are two ways of placing an 'R' on a sheet of paper, each giving rise to a different orientation of the 'R' on the paper, so too there are two ways of placing a hand in space, each giving rise to a different orientation of the hand in space.