[BUG] - Extremely slow synchronization on high end dedi

[Displee]

External

Area Other

Summary I'm using cardano-db-sync.

It took me 5 days to sync the Cardano blockchain to 99.50%. It's extremely slow on a high end dedicated server. Server specs:

- Software
   Debian 10 (Buster)
- CPU
    Intel(R) Xeon(R) CPU E3-1270 v6 @ 3.80GHz
    Cores : 8
    Cache : 8192KB 
- Memory
   2x 16384MB (32GB)
- Storage
   2x 1TB HDD (2TB)
- Internet speed
   1 GB/ps

I noticed it barely uses my CPU. It does however use some RAM. Here's a htop snapshopt:

But that's mostly the postgres service...

Here's a speedtest during synchronization:

I can't lay my finger on the issue. The logs do not contain anything outstanding either. I've restarted everything multiple times, some times after two days, some times after a day, but no effect. It's still extremely slow. Here's the latest log: https://displee.com/upload/cardano-db-sync-log.txt

My docker compose: https://displee.com/upload/db-sync-docker-compose.yml I added 2 lines to the docker compose (in the postgres service):

ports:
      - 5432:5432

Does anyone else have experienced the same issue before? Can anyone help me understand why this is going so slow?

[gitmachtl]

Try to sync with only one or two nodes in your topology, it slows down if you fill it up with like 20 nodes in there.

[gitmachtl]

@ggcaponetto please remove your passwords from your posting and also all your personal and wallet information! you're giving out way too much information here.

[ggcaponetto]

@gitmachtl thank you very much. I am ashamed.

[gitmachtl]

@gitmachtl thank you very much. I am ashamed.

no problem. but please change all the passwords and wallet addresses that were exposed in this posting.

[ggcaponetto]

Back to the issue. @Displee Maybe it's worh checking out if the presumably low IOPS of your HDD are the bottleneck.

[Jimbo4350]

Closing this. If this is still relevant please reopen.