I usually build my own kernels, and run them on an otherwise all-RedHat system. The recent failure of a GLibC update to run on anything other than a recent RH kernel brings a question to mind: why would I want to run a Red Hat kernel?You half answered yourself here: because Red Hat tests updates on a Red Hat system using Red Hat kernels.
There's the convenience factor, of course, but stable kernel releases (I don't run the testing versions) are infrequent, so not much time is spent building them.You and I and lots of others don't have problems building our own stuff. Most users do. Updating using RPM is an order of magnitude easier for non-technical users.
What are the benefits of running RH's kernels as opposed to the plain-vanilla current stable kernel?The Linus kernels get released when LT thinks they're ready. LT doesn't have to answer to lots of angry customers when things break, so the RH releases tend to have had more testing and fixes. Further, features like NPTL (backported from 2.6) are added to the stable series, so you can use the improved performance in production now instead of having to rely on a developer kernel.
Kj
-- Shrike-list mailing list Shrike-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/shrike-list
