I’ve been getting the “Raft. Ready took too long to process” on dgraph alpha and I’m noticing long delays on the server.
I’d like to figure out how to avoid this. Is there a way to see which queries or transactions are causing the error? It would greatly aid in trying to speed things up. I have no doubt that I’m doing something inefficient, but a tool to help me narrow down the problem would be wonderful.
The Raft.Ready log is a warning log that appears when Raft.Ready takes more than 200 ms. The log message shows a break down of the time taken for saving to disk, disk sync, applying proposals, and advancing Raft.
The typical culprit for this warning log are the disk latencies. Using a fast high-IOPS SSD would help reduce the disk latencies so the warning log won’t show up.