cache craziness?

can someone sanity check something for me? i'm using sst + trpc + drizzleorm + aws + rds + postgres on the server, and i guess trpc + tanstack-query on the client side... when i do mutations, to say delete something from a list and go optimistically update the client state, via useUtils setData etc... everything is fine... then a few seconds later the browser does a request (tanstack query?) and gets OLD data back from "disk cache" and updates the view to how it was before my mutation! then a few seconds later it updates again with the correct data... i'm having a hard time figuring out which cacheing layer is screwing up... has anyone ever seen such behavior?
Are you canceling your queries when you are mutating? If you are then I wish you good luck, because that sound pretty tough.

Looking for more? Join the community!