30 May
2023
30 May
'23
1:28 a.m.
On 5/29/23 20:25, Dan van der Ster wrote:
Hi,
Sorry for poking this old thread, but does this issue still persist in the 6.3 kernels?
We are running a mail cluster setup with 6.3.1 kernel and it's not giving us any performance issues. We have not upgraded our shared webhosting platform to this kernel yet (where we experienced this issue). But it does not seem like it is still an issue. A CephFS user asked this question during Cephalocon 2023, see [1], and Patrick answers that Xiubo Li got that fixed (supposedly a read ahead issue) but did not know if it was already mainlined. Gr. Stefan [1]: https://youtu.be/MRXbU96BMoE?t=1773