ord
ord copied to clipboard
ordinals
Ord indexing process extreme disk usage
I've been building an index with --index-sats enabled, using the latest ord and I set to find the main bottleneck constraining the speed since it doesn't seem to have gained much of the speed improvements that came with v0.5.1. I'm using macOS for this test, and I was shocked to see that the ord process had written and read more than the entire bitcoin blockchain worth of disk space. This struck me as extremely odd, considering that I'm at block 370000 and the resulting index.redb file is only at 13GB at the moment. Is there something terribly wrong with the way this process is done? Am I missing something obvious?
(Disk resources used by `ord` when indexing with `--index-sats` at block 370000)
0.5.0 had written a staggering 3.5 TB before it never even finished (without --index-sats). Also on mac. I wondered if it's broken. Now trying 0.5.1, let's see.
