Every GraphQL API has its own limitations and we can not query an arbitrarily large number of nodes because every query has a cost and the API will return an error if the connector exceeds the maximum cost. The query cost function is directly dependent on the page sizes. The connector uses its own algorithms for finding the page sizes but also gives you the freedom to tweak them based on your needs. There are two ways you can achieve this behavior:

• Specify custom page size weights by including the "other:pageSizeWeights" attribute. Let's suppose that we have a "Customers" table which has two nested GraphQl connections, businesses and customers. If you want a specific ratio between these two entities per query, let's say 200 customers per 1 business, you should specify "other:pageSizeWeights=1:200" in the respective script. The advantage of this attribute is that you do not bypass our logic for finding the optimal page size based on the query cost.
• Specify custom page sizes by including the "other:pageSizes" attribute. By specifying this attribute you will bypass any check we do for query cost, so you should be careful because there will be API errors if you exceed the maximum cost per query. Example: "other:pageSizeWeights=10,30", retrieve at most 10 businesses and at most 30 customers per business in one query.