Robert Moskowitz wrote: > Karanbir Singh wrote: > >> Robert Moskowitz wrote: >> >> >>> I do an rsync to my local repo. I had looked at the logs just to see >>> what came over, and saw the new kernel. Of course I can't see what the >>> xml is saying is currently available... >>> >>> So then I ran yum update on some of my systems. Some picked up the new >>> kernel one did not (so far). >>> >>> >> you seem to run into this problem often, it might be worth actually >> looking to find the part of your process that is broken and fixing it. >> >> within the CentOS setup, metadata for a package is only created once and >> only at one place and is done within a few seconds of the email being >> sent out announcing the package. >> > This really has me buffaloed. > > I suppose I can run a clean and see what happens, but the systems that > got the new kernel, got it, and how do you take it back to try again? > > And the system that that did not get, well, I suppose a clean might see > if it gets it now. Stay tuned. Yes a yum clean all on the notebook that did not get the kernel is now not only getting the kernel but also: cups and enscript. So I had some problems here. The metadata or dbcache? This is my notebook, and when I am away at conferences, I change my source of update and base to the mirrors, and when I get back I change back to my local repo. That might be why it got lost? I guess I should add a 'yum clean all' whenever I make these changes.... But I guess you did not announce the kernel, but anyone pulling the updates locally would be getting them?