Re: Question for a kernel developer "in the know" - RE: [Non-DoD Source] Re: Can't get RAID5/RAID6 NVMe randomread IOPS - AMD ROME what am I missing?????

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 23/09/2021 22:05, Finlayson, James M CIV (USA) wrote:
All,
I have a question and this is silly but I have to share.   I'm well beyond the age to now dive into kernel code to diff it, but I have a question for those "in the know".

We switched from the 5.14rc4 kernel and rc6 (I believe) to the stock 5.14.0 kernel and my >84GB/s sustained bandwidth is now hovering at 100GB/s with no changes whatsoever by me in any tuning of the mdraid or the SSDs.   My question revolves around when there are "rc" kernels, is their extra debugging tracing or any other built in "inefficiencies" in release candidates that get configured out of the released kernels?

I'm puzzled by the performance gain of > 20% when I did nothing that I can take credit for.

Bear in mind the whole point of rc is to shake out minor glitches. Linus takes a whole bunch of "improvements" into rc1, and they get shaken down over the following rcs. So I guess someone put in a load of improvements that got selected for rc1.

They may not initially have worked as well as planned, and other people would have spotted problems. I get the impression there's a lot of room for improvement, and I think Christoph Hellwig has been working hard on this. So there's a good chance some of his stuff got into rc1, and he's kept tweaking it through the next rcs. And you've benefited!

Cheers,
Wol



[Index of Archives]     [Linux RAID Wiki]     [ATA RAID]     [Linux SCSI Target Infrastructure]     [Linux Block]     [Linux IDE]     [Linux SCSI]     [Linux Hams]     [Device Mapper]     [Device Mapper Cryptographics]     [Kernel]     [Linux Admin]     [Linux Net]     [GFS]     [RPM]     [git]     [Yosemite Forum]


  Powered by Linux