We hope to fix the problem permanently in the next 2-3 hours, as our hosting company now moves the faulty Elasticsearch virtual machine to another hardware node that is less crowded with other high I/O intensive neighbours.
We apologize for the delays in trace processing.
Edit 14:57: The delays are now piling up in advance of the node migration and we see around 45 minutes of delay in traces. If you urgently need access to a trace, as a workaround the delay, if you use the tideways CLI or chrome extension, those traces are prioritized higher for processing.
Edit 15:56: The node migration was successful and our trace backlog has been cleared. Everything is caught up and operational again.