I've been running geth for about 2.5 days so far, and I'm trying to work out why it's taking so long to sync.

It's downloaded ~40Gb of the blockchain (up to block #2379173 if I'm interpreting the output correctly) and is running at consistently <25% CPU, often less. The IO load varies a lot but is on the order of 5-10MB/s, if that was purely for downloading and writing the blockchain data, it would take about half a day, rough estimate. This implies that the IO load is not for writing the chain to the disk. The network load is tens of KB/s, and my connection can handle a lot more than that. This means there's some MB/s of IO (writing) which is neither network chatter nor a net disk write.

The bottleneck clearly isn't the CPU, and the these numbers don't support IO as the bottleneck either. So why is it taking so long?

  • what is command line you used to start geth? Commented Jun 19, 2017 at 12:19

1 Answer 1


Is your disk a spinning, magnetic disk (as opposed to an SSD)? If so, your disk might be the bottleneck: syncing the blockchain involves a lot of random reads and writes. If your OS supports it, you can check your disk I/O depth queue and/or percent utilization.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.