I've observed a consistent increase in the "Swap used" metric, reaching levels at least 50% usage after extended periods of uptime on an OrangePi 5 Plus with 16GB of RAM. Despite small fluctuations, the trend is consistently upward, necessitating a system reboot to restore swap levels to reasonable values.
After 2.3 hours of uptime:
After 5 days of uptime:
In the following graph, the only points where swap is close to 0% is after a reboot. After reboot, it gradually increases again:
On an OrangePi 5 Plus with 16GB of RAM.
Running:
geth v1.13.4
lighthouse v4.5.0
Even after a reboot, the upward trend resumes.
Is there a memory leak causing this increase in swap usage?
This looks to be a hardware issue. 16GB RAM is just not enough. I have the same image and config running on an OrangePi 5 Plus with 32GB of RAM and the swap has never gone over 1%.
I've observed a consistent increase in the "Swap used" metric, reaching levels at least 50% usage after extended periods of uptime on an OrangePi 5 Plus with 16GB of RAM. Despite small fluctuations, the trend is consistently upward, necessitating a system reboot to restore swap levels to reasonable values.
After 2.3 hours of uptime:![2-3-hours](https://github.com/EOA-Blockchain-Labs/ethereumonarm/assets/6164653/c7530de3-1636-48b6-b733-4daf0e16cc41)
After 5 days of uptime:![5-3-days](https://github.com/EOA-Blockchain-Labs/ethereumonarm/assets/6164653/81e20fa8-301b-48d4-8dfb-bac9d247d91f)
In the following graph, the only points where swap is close to 0% is after a reboot. After reboot, it gradually increases again:
On an OrangePi 5 Plus with 16GB of RAM. Running: geth v1.13.4 lighthouse v4.5.0
Even after a reboot, the upward trend resumes.
Is there a memory leak causing this increase in swap usage?