fix: optimize the round trips when reconciling the subgraph deployments

[abarmat]

Issue

I was running the Indexer Agent and after a few days I realized the amount of queries it sent to the network subgraphs was out of proportion and quickly draining my api key.

Investigation

The Indexer Agent performs a number of “housekeeping” queries at regular periods of time to sync with the latest network status.

One of those queries gets the entire set of subgraphs on the network but it does so paginating by 10 items batches. With the current number of subgraphs over 10,000, it takes more than 1,000 queries to get them all.

The eventual runs every 120 seconds, that means it's sending about 720,000 queries a day.

Resolution

Set the first parameter to 1000 for the query within subgraphDeployments function of the NetworkMonitor.

[dwerner]

Good find. I have another PR in that was intended to fix a test in that same function (#1056), I bumped the page size there as well.

[fordN]

🙌

[dwerner]

closing in favor of #1056, thanks!

[abarmat]

@dwerner it seems that #1056 is not including this change, is that right? take a look at this line

indexer/packages/indexer-common/src/indexer-management/monitor.ts

Line 648 in 7acb873

first: 10,

[abarmat]

This PR is fixing a different function than #1056, this one is for SubgraphDeployments and the other one for Subgraphs.

[abarmat]

Also tagging @fordN for visibility

[dwerner]

My bad, I mixed up a similar fix.