Querying stakepools with limit=1000 only returns 100 results

[rcmorano]

Environment

• https://graphql-api.testnet.dandelion.link/
• https://explorer.cardano-testnet.iohkdev.io/graphql/

On my side (dandelion endpoint), I'm using these official docker images:

• inputoutput/cardano-graphql:2.1.0
• inputoutput/cardano-graphql-hasura:2.1.0
• inputoutput/cardano-db-sync:5.0.1

Platform

• Linux / Ubuntu
• Linux / Other
• macOS
• Windows

Platform version: Ubuntu 20.04.1 LTS

Runtime

• Node.js
• Kubernetes, Docker

Runtime version:

/application/packages/server/dist # node -v
v12.15.0
/application/packages/server/dist # 

Kubernetes/Docker:

$ kubectl get nodes -o wide
NAME                                             STATUS   ROLES    AGE   VERSION   INTERNAL-IP     EXTERNAL-IP       OS-IMAGE                        KERNEL-VERSION     CONTAINER-RUNTIME
_REDACTED_  Ready    <none>   36d   v1.18.3   10.69.132.45    _REDACTED_       Ubuntu 20.04.1 LTS a25c031bef   5.4.0-42-generic   docker://19.3.12

Steps to reproduce the bug

1.- Query (using official example query) one of the provided endpoints with limit=1000 query variable:

# my instance
curl -s 'https://graphql-api.testnet.dandelion.link/' -H 'Accept-Encoding: gzip, deflate, br' -H 'Content-Type: application/json' -H 'Accept: application/json' -H 'Connection: keep-alive' -H 'DNT: 1' -H 'Origin: https://graphql-api.testnet.dandelion.link' --data-binary '{"query":"query allStakePoolFields (\n    $limit: Int!\n)  {\n    stakePools (limit: $limit, offset: 100) {\n        fixedCost\n        hash\n        margin\n        metadataHash\n        owners {\n            hash\n        }\n        pledge\n        relays {\n            ipv4\n            ipv6\n            dnsName\n            dnsSrvName\n        }\n        retirements {\n            announcedIn {\n                hash\n            }\n            inEffectFrom\n        }\n        rewardAddress\n        updatedIn {\n            hash\n        }\n        url\n    }\n}","variables":{"limit":1000}}' --compressed | jq '.data.stakePools | length'
# official explorer instance
curl -s 'https://explorer.cardano-testnet.iohkdev.io/graphql' -H 'Accept-Encoding: gzip, deflate, br' -H 'Content-Type: application/json' -H 'Accept: application/json' -H 'Connection: keep-alive' -H 'DNT: 1' -H 'Origin: https://explorer.cardano-testnet.iohkdev.io' --data-binary '{"query":"query allStakePoolFields (\n    $limit: Int!\n)  {\n    stakePools (limit: $limit, offset: 100) {\n        fixedCost\n        hash\n        margin\n        metadataHash\n        owners {\n            hash\n        }\n        pledge\n        relays {\n            ipv4\n            ipv6\n            dnsName\n            dnsSrvName\n        }\n        retirements {\n            announcedIn {\n                hash\n            }\n            inEffectFrom\n        }\n        rewardAddress\n        updatedIn {\n            hash\n        }\n        url\n    }\n}","variables":{"limit":1000}}' --compressed | jq '.data.stakePools | length'

2.- Query with limit=100 and offset=100 returns the rest of the stakepools confirming they are there and first query limit just doesn't work:

# my instance
curl -s 'https://graphql-api.testnet.dandelion.link/' -H 'Accept-Encoding: gzip, deflate, br' -H 'Content-Type: application/json' -H 'Accept: application/json' -H 'Connection: keep-alive' -H 'DNT: 1' -H 'Origin: https://graphql-api.testnet.dandelion.link' --data-binary '{"query":"query allStakePoolFields (\n    $limit: Int!\n)  {\n    stakePools (limit: $limit, offset: 100) {\n        fixedCost\n        hash\n        margin\n        metadataHash\n        owners {\n            hash\n        }\n        pledge\n        relays {\n            ipv4\n            ipv6\n            dnsName\n            dnsSrvName\n        }\n        retirements {\n            announcedIn {\n                hash\n            }\n            inEffectFrom\n        }\n        rewardAddress\n        updatedIn {\n            hash\n        }\n        url\n    }\n}","variables":{"limit":100, "offset":100}}' --compressed | jq '.data.stakePools | length'
# official explorer instance
curl -s 'https://explorer.cardano-testnet.iohkdev.io/graphql' -H 'Accept-Encoding: gzip, deflate, br' -H 'Content-Type: application/json' -H 'Accept: application/json' -H 'Connection: keep-alive' -H 'DNT: 1' -H 'Origin: https://explorer.cardano-testnet.iohkdev.io' --data-binary '{"query":"query allStakePoolFields (\n    $limit: Int!\n)  {\n    stakePools (limit: $limit, offset: 100) {\n        fixedCost\n        hash\n        margin\n        metadataHash\n        owners {\n            hash\n        }\n        pledge\n        relays {\n            ipv4\n            ipv6\n            dnsName\n            dnsSrvName\n        }\n        retirements {\n            announcedIn {\n                hash\n            }\n            inEffectFrom\n        }\n        rewardAddress\n        updatedIn {\n            hash\n        }\n        url\n    }\n}","variables":{"limit":100, "offset":100}}' --compressed | jq '.data.stakePools | length'

What is the expected behavior?

Paging works as expected and return more than 100 results.

[rhyslbw]

Thanks for raising this again @rcmorano. The problem here is that we have a static limit in Hasura as a easy way to protect the API from abuse, but clearly introduces misleading results as you have documented. I will revisit this before the next release