Custom Query (894 matches)

Fix case where packages are listed in the configuration multiple times show up as incorrect, even when correct.

In 0.8.0, when I run a client and tell it to remove rpms, it gets confused and claims it hasn't removed anything:

...
Redhat[verbose]: Phase: final
Redhat[verbose]: Correct entries:       538
Redhat[verbose]: Incorrect entries:     0
Redhat[verbose]: Total managed entries: 538
Redhat[verbose]: Unmanaged entries:     4

However, the packages really are gone.

Here's a ticket to talk about the "growing memory use" problem. Using 0.8.0, I just ran the client on 62 nodes and saw this change in server memory usage:

Before:

root      2108  0.1  5.2 115444 108192 pts/11 S   Feb02   6:59 /usr/bin/python2.3 /usr/sbin/bcfg2-server

After:

root      2108  0.1  5.6 123176 115956 pts/11 S   Feb02   7:36 /usr/bin/python2.3 /usr/sbin/bcfg2-server

The only thing done between these data points is running the 62 clients. You'll also note that the beginning memory size is already quite a lot larger than the ~46MB we saw it using last week.