[Bug] Sorting subgraphs by nested FDS field throws error

Bug report

Querying the graph-network arbitrum subgraph and sorting by metadata__displayName; throws an error (see log output).
There was a recent change to the network subgraph to support File Data Services; this change moved the displayName field from the top-level Subgraph entity to be in a nested SubgraphMetadata entity that is controlled by FDS.
Sorting by any field in the metadata entity throws this error.

Relevant log output

Results in this error message from graph-node


message: `Failed to get entities from store: column cc.block_range does not exist, query = /* controller='filter',application='sgd675036',route='d8217ff92881cd2a-9ef95e8b5fa290fd',action='174820478' */\nselect 'Subgraph' as entity, to_jsonb(c.*) as data from (select c.*\n from "sgd675036"."subgraph" c left join "sgd675036"."subgraph_metadata" as cc on (cc."id" = c."metadata" and cc."block_range" @> $1) \n where c.block_range @> $2 and (c."active" = $3 and c."entity_version" = $4 and exists (select 1 from "sgd675036"."subgraph_metadata" as i where i."id" = c."metadata" and i.block$ <= $5 and (left("display_name", 256) != left($6, 256))))\n\n order by cc."display_name" asc, c."id" asc\n limit 20) c -- binds: [174820478, 174820478, true, 2, 174820478, ""]

IPFS hash

No response

Subgraph name or link to explorer

https://api.thegraph.com/subgraphs/name/graphprotocol/graph-network-arbitrum

Some information to help us out

• Tick this box if this bug is caused by a regression found in the latest release.
• Tick this box if this bug is specific to the hosted service.
• I have searched the issue tracker to make sure this issue is not a duplicate.

OS information

None

Inizia riproducendo la query GraphQL collegata rispetto al subgraph arbitrum di graph-network e ispeziona la query dello store di graph-node mostrata nel log, in particolare il join a subgraph_metadata e la relativa condizione block_range. Traccia come viene generato l’ordinamento per campi di metadati annidati; il lavoro è completato quando l’ordinamento per campi annidati non produce più l’errore di colonna mancante e la query segnalata restituisce risultati.