Review GraphQL endpoints to avoid node overload

[xgreenx]

Some GraphQL endpoints have complexity O(n), O(n^2), where n can be a number of arguments or internal data stored in the blockchain.

In the mainnet, it can affect GraphQL nodes and cause them to be stuck.

Each request should have some limitations. It can be a limitation of input or output arguments or query frequency.

The code has many simple queries (like getting block by height, transaction by id, and so on). The DevOps and infrastructure can limit the rates for those requests(for example, Apollo, but it requires investigation). So we don't need to modify the code and only care for it during cluster deployment.

We need to review all queries and decide which queries should be limited in the code(we need to implement some custom logic, maybe pagination, etc.) and which by the infra team(also, we need to decide how to do that).

[ControlCplusControlV]

GraphQL endpoints should probably also implement some size check on request bodies to prevent that DOS vector