<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix">On 24/03/2023 14:52, Björn Ekelund via
Dxspider-support wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CABmD_KRpDf66HedDX9EG8aKy7EYBMjFzyn_YZbB1qux0+p0w5Q@mail.gmail.com">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<div dir="ltr">Thanks Kin,
<div><br>
</div>
<div>the debug subfolders are gigantic. Is there a way to prune
this data automatically?</div>
<div><br>
</div>
<div>Björn</div>
</div>
<br>
</blockquote>
<br>
First thing, if you are not interested in the full panoply of debug
info available then, in a console type: set/deb progress nologchan<br>
<br>
That by itself will reduce the debugging to less than 10% of the
full output. <br>
<br>
To make it 'stick', edit /spider/local/DXVars.pm and add those debug
levels above into @debug = (chan state ....)<br>
<br>
The standard issue mojo version looks like this (near the bottom of
DXVars.pm):<br>
<br>
<blockquote><font face="monospace"># are we debugging ?<br>
@debug = qw(chan state msg cron connect progress nologchan);</font><br>
</blockquote>
<br>
Which should do just fine for a "normal" node.<br>
<br>
And the second, by default, the node will prune the debugging down
to the latest $DXDebug::keepdays of files (normally 10) by itself,
but you can change to something a bit smaller (probably no less than
5), if you like, in the startup script. Actually, thinking about it,
you could just do 'set/debug progress nologchan' in that startup
script while you are in there. <br>
<br>
@debug = (...); was the original way of doing this, so I still do
that. I am old-fashioned that way :-)<br>
<br>
Dirk G1TLH<br>
</body>
</html>