I'm syncing a geth archival node and the syncing is becoming progressively slower. Current sync status is block ~11.5M/12.5M and syncing 1 block every 2 seconds.
Command I'm running is geth --syncmode full --gcmode archive --txlookuplimit=0 --cache.preimages --datadir /datadrive --cache 128000 --snapshot=false
I'm running my node on Azure with a Standard E16as_v4 (16 vcpus, 128 GiB memory) instance and 16TB premium SSD storage attached disk. OS is Linux (ubuntu 18.04).
I am struggling to understand what the bottleneck is. I have tried connecting to new peers from https://etherscan.io/nodetracker, optimizing caching, running on a faster disk, and running with more RAM.
How do I speed up this sync or identify the bottleneck?