Parler appears to have had an inefficient architecture that required an absurd amount of compute power to run compared to their customer base. Apparently they w

Parler appears to have had an inefficient architecture that required an absurd amount of compute power to run compared to their customer base. Apparently they were running one hundred metal instances for their database alone; speculation on Twitter was that they’re struggling to find someone who can give them that much capacity in a hurry.