Some feedback from a Ovirt deployment with standard 4.14 kernel from kernel.org for over 9 months. This is no recommendation to go that way but a personal opinion. First of all Redhat kernels are very stable and the use is very convenient. But we always disliked the fact, that Redhat/CentOS kernels are branded 3.10.x while they are a total mixup of patches and enhancements. Some examples why both ways have up and downsides. 1st example for CentOS being late with patches: https://www.mail-archive.com/kvm@vger.kernel.org/msg124576.html added patches targeted at 4.5 for Windows migration bugs. These are in 4.14 mainline since its first release in November 2017. https://git.centos.org/blob/rpms!kernel.git/fd768ebf7d64cd053f749eae78e72... added the same patches to the old 3.10 base kernel a year later in December 2018. 2nd example for CentOS enhancements: CentOS kernel 3.10 provides blk-mq support although that feature was added in 3.13 and later mainline kernels 3rd example for instable kernel.org versions: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?... gives a funny insight about problems during kernel development and backports. That resulted in unstable versions from 4.14.94 to 4.14.96. To sum it up. Mainline longterm stable feels more dedicated to the policy to add only fixes and no enhancements. Regarding stability had no problems yet and no case where we missed patches from the CentOS tree If you like to try it out, I uploaded a spec and config that I once took from elrepo.org. Just add the kernel.org 4.14 tgz source and the two missing elrepo cpupower files. You should be ready to go with a simple "rpmbuild kernel-lt-4.14.spec". config-4.14.97-x86_64: https://ufile.io/n1og8 kernel-lt-4.14.spec: https://ufile.io/ogp12 We will jump back onto the standard bandwagon when going to CentOS 8 with kernel 4.18. Best regards. Markus