Hi @doingitnow
Internally we use NATS for logs.
What happened is that the logs cluster might have been clogged and sent batches at those moments you were looking at. Definitely not great but we are looking into improving that, it’s not an issue on your end at all.