Description of problem: So I'm sitting at home, and my not-even-on-the-net laptop squawks that it's just received new e-mail. Turns out it's yum-updatesd telling me there's a problem updating the system. True - several RPMs can't upgrade because it includes an update to a shared lib, and several *other* RPMs haven't dropped updated versions into Rawhide yet, and still need the *old* libraries. But there's no reason to send out 27 e-mails, once per hour. Nothing has changed since the last time it nagged me. Recommended change - do a stat() on the involved yum database files and the RPM databases - if none have an mtime less than $run_interval seconds ago, then nothing has changed, so don't bother telling the admin *again*. Oh, and probably should include /etc/yum/yum-updatesd.conf in the list of files to stat, so it DTRT even on the first iteration. Version-Release number of selected component (if applicable): yum-3.2.20-6.fc11.noarch How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info:
The checking (and thus the email) is only done at run_interval. The default run interval is once an hour though.
IMO it *is* a bug - there is a difference between the "check interval" (1 hour is reasonable) and a "renotification" interval (1 hour is definitely unreasonable for email). Currently yum-updatesd lacks the much needed capability to differentiate between the two. Definitely a valid RFE - which I previously reported as bug 458311 (closed wontfix as well), tried reporting through the RHEL support channel, but somehow can't get anyone to agree it's a valid problem.