S1 Pod fullness at 100% with only 183k vector count

I am using 1 S1 pod to upsert my 1024 dimension embeddings. According to the documentation, we can upload around 4 million vectors in S1 pod of 1024 dimension.

Why does my pod fullness go to 100% only with 183k vectors.

Hi @p22saurabhs. There are other factors that effect fullness than just vector count. That’s why we stress in the docs that anything referencing total count numbers is just a guideline.

The last fullness report I see for your index it was less than 1%. Can you share the screenshot where you saw your index was at 100% full? Also, do you see the same result when you run describe_index_stats() on the index?

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.