I’m the (recent) proud lessee of 2 IWorx installations, one older – probably from around the 1.7 era, is Redhat 9 and in a remote DataCenter, one newer 2.1.1 era, CentOS 4.2 and sitting next to me. A few differences that I’ve noticed:
The LogWatch between the boxes are not only different versions (which I would expect, considering RedHat’s End of Life), but log quite differently in terms of their verbosity. The older installation (via email) is only logging a) Kernel b) proFTP and c) vpopmail (today). The new installation (via email) is logging a) ClamAV b) httpd c) init d) Kernel e) pam_unix f) SSHD and g) Disk Space.
I realize that some services only log if they are used that particular day, but in the case of Disk Space – that’s been missing from my LogWatch for quite some time and I’ve never seen the ClamAV section, though I have the service installed and running.
Clustering and NFS show up as menu options on the new box and are missing completely from the old. This may be you aren’t offering certain services inside particular DataCenters, (though, why a DC wouldn’t want you to potentially rent more servers from them for clustering, etc. I’m not sure) … just looking for clarification on this one.
This is to be expected, you need CentOS 4 (with NAT installed and the inteworx and vpopmail directories moved outside /home to new locations), or newer for clustering and those menus don’t show up on a box that doesn’t support clustering.
You need CentOS 4, Remember CentOS 4 is basically the same as RHEL 4, and who knows it might work on RHEL4 as well if NAT was installed and configured and the other things it mentions are done. That’d really be an answer for Chris.
As for the other, the logwatch config file allows you to specify exactly what you want it to show. I suspect if you replaced the logwatch.conf file on the old box withthe one from the new one it’d be just as verbose.
I have two logwatch.conf files. Anybody know the difference?
I suspect if you replaced the logwatch.conf file on the old box withthe one from the new one it’d be just as verbose.
… though, I thought there was some caveat for a service having it’s logging “toggle” set at the time of compile (which is why I thought most of us have a pile of unwanted vpopmail entries that are seemingly unstoppable (?) ).
It feels more like the logging options of the various services have had their logging switches ‘reset/overridden/etc’ along the way on the older box.
I’ll check my actual LogWatch file and see how it compares to yours shortly.
Thanks again,
JB
EDIT I checked the LogWatch.conf on my older RedHat box and the only difference I saw between what you’ve posted and mine was the level of detal, yours being “High” and mine being “Low.” However both of my boxes in question are set to “Low,” with “Services=All” and produce different output (again, no ClamAV or Disk Space on the older box, etc).
… Pascal is right that the logging is a compile time option and not configurable after the fact…
… which, as the Linux neophyte that I am makes me think all (most/many?) services must have that option, and maybe for some reason over the course of updates, certain services have had their switch to log blown-out. If no one else is noticing this (which would be good), then it’s isolated to me and not a bug at all (also good). It doesn’t tell me why I have the issue, but at least takes it out of the bug category and into something else.
After watching more carefully the last couple of days of logs, it looks like the old RedHat box is not logging:
httpd
crond (under Pam - but does log su)
disk space