[Pan-users] Re: Reproducible crash retrieving headers in million-header group

[Wayne E . Nail]

Zing <address@hidden> writes:

> 
> On Fri, 30 Jul 2010 14:57:09 +0000, Wayne E. Nail wrote:
> 
> > Here's a wrinkle though; I run Pan in stunnel with news server set to
> > localhost:119 and the secure NSP url stored in
> > /etc/stunnel/stunnel.conf. Might this setup be hitting a memory wall Pan
> > by itself would not?
> 
> Well, if you run "top" and hit "M", the top memory eaters should be 
> listed...  but could this be selinux?  run "setenforce 0" as root and try 
> again.
> 

selinux!!! There is no circle of hell yet wrought to torture selinux as it
deserves! Not only does it thwart everything any reasonable person would ever
want to do, it *doesn't even tell us it has done*! <gnash><rend><smash><sob>

<wiping of mouth froth, combing of hair, straightening of clothing>

Actually I disable selinux at kernel level with 'selinux=0'

But, good call.

Top 'm' always shows Pan at about 32% when pulling headers in the absence of
anything else on the system wanting more; gkrellm is merely a lazier way to see
the same information within a wider beam. Essentially I have found Pan an app
that plays well with others.

As further proof that the definition of insanity as 'doing the same thing over
and over again and expecting different results' failed to take into account
those who deal with computers (and especially WinOS), I am running the exact
same test under gdb once again, after having installed all the requested debug
packages. I don't really expect anything useful as a result, but you never know.

_________________________

Duncan: Actually I was looking for a way to avoid gmane's complaints about 80+
character lines *when I use the gmane dialog to compose a message*. The other
stuff... as usual you taught me things, but they didn't address the question. I
thought from past threads that you used gmane to manage pan-users dialog.