'state entries over 1 billion during BSC fullnode sync
We've been trying with a friend to run our own BSC full node on a VPS, in order to run scripts on it to be able to monitor as fast as possible arbitrage and scalp opportunities. We took a VPS from Kamatera running on Ubuntu, pretty solid stuff, 16c CPU, 64GB of RAM, 2TB SSD Storage, very solid internet connection. I don't have infos about the IOPS, but i ran a benchmark on a small VPS i had from Kamatera and it went over 18k IOPS, so both should be pretty much the same.
Our 1st attempt failed, something got corrupted during sync and Geth did not even want to start anymore and was throwing us an error, so we decided to erase everything and start a new fresh sync.
The block download phase lasted the night, maybe 12-15 hours, and now again we are at this "state entries" stuff going on, which started about 30 hours ago.
Even though it is pretty hard to understand what that thing is (seems nobody knows what the heck it is about), it feels something is wrong. I see people mentionning that they went up to 300M states entries, sometimes 500M and getting synced in like 1 day. We have currently over 1 billion of them, our last sync before it got corrupted got close to 2B.
When i check the panel on Kamatera, i see that nothing is in pain about the server, CPU is about 15% constantly, RAM is about 50% used all the time, and IOPS on the SSD are usually cruising at 2k IOPS and sometimes peaking at 3.5k or 4k before coming down to 2k again.
I have about 980Gb worth of Geth stuffs on the SSD, increasing slowly.
Is there something i am missing ? I add the picture of geth sync and also the command line that started geth on server reboot. Thanks everyone for your help, everything will be appreciated
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|