Search results

Thank you, Eric. That all makes sense. I wonder why I didn't come across this problem with other arrays, maybe I had atime off.

I have a backup disk with one ZFS pool and a number of filesystems that should receive incremental snapshots. For some but not all filesystems it refuses with "cannot receive incremental stream: destination ... has been modified". I am sure that I did no such modification, this is purely a...

But most of that software is actually written in CUDA, Eve if ROCm was a complete competitor it (which it isn't) you still don't have that software ported. AMD also put a couple of gotchas such as ROCm only working on a very small collection of cards, whereas CUDA runs on pretty much any...

Do memtest next. Personally I don't see a problem with 90 C.

What was it?

Hm. It seems that you have to build the port until the packages are upgraded.

The x server port is at 21.1.12. So a simple update should do.

There is a difference between starting to use swap and continuing to use swap. You start using swap when memory is moderately low-ish. But you don't put those pages back into RAM just because RAM becomes available. To find out what is currently in swap kill -9 random things. If it was...

LOL

With compression on ZFS will be a lot faster for compressible things like the base OS.

My hunch is that ZFS itself is heavyweight and if the machine has difficulty coping then compression or not doesn't make much more of a difference.

Lower kernel hz to 100.

Why do you think systemd could work?

The power level and limits are only in newer generations of CPUs. Use i7z to find out which frequency you actually run at.

I hear that this is actually a general mechanism that preloads disk locations that were in heavy use before. I.e. not hardcoded to specific applications. Whether that's true seems to be hard to google.

How do you measure current memory consumption?

But you bought the slowest disk on the market...

You have general instability, with no exact reproducibility. It is not possible to predict with a guarantee what will make it go away. But given that you have a CPU cooler too small for your high-end CPU that is clearly the first thing to try. I don't believe that those assumptions about...

The thermal throttling is obviously not working as advertised, otherwise you wouldn't have errors. That is why it is kind of hopeless to chase a stable underclock, which effectively is what you do. The temperature in your room goes up a degree, and you are unstable again. Also, I think your...

The panics are not always the same, I think that was established on page 1 or 2. As far as the memory errors are concerned, they aren't necessarily caused by bad RAM cells. I think it is more likely the CPU flipping some bits. Again, a strong hint here is that the specific errors vary.