Since the beginning I’ve had problems with packages (either LND or Nethermind) causing “out of memory” kernel errors and high CPU usage. I created my VM with 12 GB RAM and I run following packages:
Recently I’ve increased amount of RAM to 16 GB and for the last 2 days CPU isn’t so much occupied and I don’t see any errors.
Remaining question is: what’s an average RAM requirement for dappnode and a few packages?
What I’d suggest here is having a look in DMS package to get more information on what is causing the high memory consumption.
You can do it manually if you go to DMS > Graphana > DAppNode Monitor or you can open this link. Then, check your DAppNodePackage Memory Usage Dashboard and report it if you find something’s using more RAM that expected.
Regarding your question it depends a lot on how you use your DAppNode or even what package you choose to use for any given purpose, it’s not the same to use turbo-get than goerli, for instance.
Even though we consider that it is possible to run ETH1.x full node, along with ETH2.0 beacon chain and a validator client with no issues using 8GB of RAM memory.
I’ve ended up increasing RAM to 16 GB. I couldn’t find any abnormalities since I can’t figure out how big RAM usage is “normal”… I can live with that.
But now my concern are:
ever growing Nethermind database: now is 530 GB and keeps growing. How big should it be currently?
Lightning and medal beacon chain CPU usage. Every so often it starts using quite a lot of CPU: