Possible bug with QDNS admin and logging

Re: Possible bug with QDNS admin and logging

From:

Date:

Time:

At 9:56 PM -0700 10/27/04, Scott Haneda wrote:
>on 10/27/04 11:47 AM, Men & Mice Support at cbuxton@menandmice.com wrote:
>
>> This is a behavior of BIND 9. If you specify a max log file size,
>> then every time you tell it to reconfig (which happens anytime you
>> change the server options), it starts a new log file. If you have it
>> keeping multiple copies of the log file, you can find the old one as
>> quickdns.log.0.
>
> I don't have a logging max size set, is there any way to get this behavior
>to stop?

I don't think you can. It's hard-coded into BIND 9.

Note that if you have the log window open in QuickDNS Manager when it
does this, you won't see new log entries until you close and re-open
the window (after which you won't see the old ones anymore).

>Also, whats the best course to get rid of these log lines?
>Oct 27 21:55:10.966 config: error: /var/named/conf/logging:33: unknown
>logging category 'statistics' ignored

There are a bunch of BIND 8 logging categories listed in your logging
file, /var/named/conf/logging. You can go in and remove them. (If you
don't want to figure out which ones, you can simply remove all the
category lines in the logging file, then open the server's Options
window in QuickDNS Manager and turn on whatever logging categories
you want.)

Logging is complicated in BIND. You can have multiple logging
channels (QuickDNS by default sets up two), and each channel has what
is basically a completely independent set of categories, severity,
etc. When an even happens in named, it checks whether any channels
want to log it, and then logs it or not as indicated. So a given
event can be logged multiple times, to different channels.

Each log message has a category and a severity. Severity ranges from
"Debug" and "Info" at the low end to "Error" and "Critical" at the
high end. "Debug" level is further divided numerically between 0 and
100.

QuickDNS lets you manage all of this. In the logging settings in a
server's options window, the first thing to notice is the channel
selector, at the top. This allows you to choose which channel you're
looking at - it isn't a way to choose which channel is actually in
use by the DNS service, since they're all in use at the same time.

For any given channel, you can choose what categories to log and what
severity to restrict it to. If you set the severity to "info", then
every message of "info" or higher severity (including "notice",
"warning", "error", and "critical") will be logged, assuming its
category is selected.

Chris Buxton
Men & Mice - Making DNS Easy
Customer Service and Sales Engineer

Sales:
888/4-DPOINT

Support:
858/452-3696