Is reporting on ECC currently implemented in pve7? If not is it on the roadmap? and if so could we have an indication by what version/release date?
I have found a few interesting atricles. I'll share one;
https://www.admin-magazine.com/HPC/Articles/Memory-Errors
I am not sure but I also think I have read that edac tools or MCE related stuff has been superceded by rasdeamon by now. I could be wildly mistaken though.
point being is that having ECC memory is of no use if one does not know that there are issues being reported to the IPMI platform or to the OS, but not reported to the root user by the PVE stack.
And please believe me, even server motherboards can have faulty IPMI implementation and not pick up on ECC errors.
If you want to go down that rabbit hole then here you can;
https://www.truenas.com/community/threads/the-usefulness-of-ecc-if-we-cant-assess-its-working.83580/
I have found a few interesting atricles. I'll share one;
https://www.admin-magazine.com/HPC/Articles/Memory-Errors
I am not sure but I also think I have read that edac tools or MCE related stuff has been superceded by rasdeamon by now. I could be wildly mistaken though.
point being is that having ECC memory is of no use if one does not know that there are issues being reported to the IPMI platform or to the OS, but not reported to the root user by the PVE stack.
And please believe me, even server motherboards can have faulty IPMI implementation and not pick up on ECC errors.
If you want to go down that rabbit hole then here you can;
https://www.truenas.com/community/threads/the-usefulness-of-ecc-if-we-cant-assess-its-working.83580/
Last edited:
