Hey! Our team (Parallelia x Thinking Machines) is trying to develop a query analytics tool that is useful and needed by our indexers.
We want to reveal query patterns and index usage analystics to equip indexers with the info to optimize their infras.
The Graph posted this RFP and we’re in talks to work on it.
Since indexing is a heavy task and query time and performance are key indicators for indexers, our idea is to develop a query analytics tool that would extract data from the Graph Node logs about each GraphQL query.
The pipeline should summarize the logs and be able to show real-time performance metrics, and analyze query patterns, and index usage analytics in a dashboard
Also, given that it has been around 6 months now since the RFP was published, we want to make sure that we work on something that is still relevant to indexers.
If this idea no longer seems relevant for you, would you have any idea for us to improve on a different set of problems which would be better to spend the efforts on?
Please feel fee to write +1/-1 in the comments or on this form if you prefer anonymity or would like to provide feedback in depth.
Definitely something Indexers would be interested and also the area most Indexers are least experienced in.
I’ve done some work myself via scripting with qlog to pull data out of the query logs and try and bring some insight to what my query nodes are serving, but its super basic bash scripting and totally neglected amongst a sea of other priorities.
Even basic stuff like queries per second over time, total queries served since inception/over a time period, longest running queries, query frequency etc. - all very valuable and I would love to be able to browse and share with other Indexers. Going beyond that into the territory of trend analysis etc sounds awesome!
I’d be happy to provide feedback on anything you are working on. @chris it might be useful to gauge enthusiasm for this type of tool via IoH?
Definitely agree with @cryptovestor that this is something Indexers would find valuable.
@aircandy I’ve sent you a DM to connect over Discord. It’d be great to set some time aside in the next Indexer Office Hours to hear more about your proposal and to collect feedback and rally support from Indexers.
Excited to announce that representatives from Parallelia x Thinking Machines will be joining IOH 50 on Tuesday 1st March to share more about their proposal and collect feedback from the community.