Since my Pi fails to ever mint any nushares, I started to mint them on my regular PC. After 12 hours of minting nu has reached a state where it consumes ~700 MiB of RAM while bitcoin-qt only consumes 415 MiB and it is being stress tested at this moment. So clearly something is badly wrong with nu. Both of those processes are running at very low priority, if that makes any difference. I remember that when I was not minting then Nu didn’t consume that much memory. The memory usage goes up after a day of minting. No wonder the Pi can’t handle it if it only has 1 GiB of RAM.
Perhaps I am old-fashioned but in my opinion such a program should not consume more than 100 MiB of RAM, ever. I wouldn’t be complaining but since it’s my PC I hate it when it gets slow and I cannot browse the Internet normally. So, naturally I will try to debug and see what process is making my system so slow and I suspect that in this case it is the fault of nubits. I hear my HDD grinding quite often and during that time everything just hangs and I am 100% sure that block chains are causing this halting and grinding sounds.
So, the question is — is anyone else having similar problems? Is it probable that in the future the code gets optimized? Or perhaps the developers have become lazy and I am expected to just buy a faster machine?
That’s actually a good point. I never thought about that possibility. If we’re both using Pi 2 and both minting blockshares and nushares, then the only difference is the number of shares we hold and the OS we use. My minting address is in top 100 if that is enough information for a meaningful comparison. However, blockexplorer.nu seems to be down so I can’t access the top 100 at the moment, who knows what has happened since I last checked it.
You can mint with an empty wallet and a wallet with one of your addresses to sort of fit a linear relation. Top 100 addresses are not good to use as many split their holdings in multiple addresses. If you are in the top 1/3 in the poll your minter will need lots of RAM…
Here’s the screenshot. It might not be a good time to measure the CPU and RAM usage since nud is still downloading the blockchain. However, it’s interesting to see that some mount.ntfs process consumes quite a lot of resources. Is it because the external USB drive is NTFS and that if it was EXT4 there would be no such overhead?