kernel panic and postfix problems

Anthony Philipp philipp1 at itg.uiuc.edu
Wed Jun 2 22:20:33 PDT 2004


ok, sorry for the late response. it (the kernel panic) happened again 
twice! i was backing up my friends laptop with rsync all three times the 
kernel panic occured. this leads me to believe that rsync is the cause of 
this panic. everytime swi1:net is the cause of the problem, only the stack 
pointer and some other pointers are different for the output. so someone 
who know more about kernel panics than me, why would rsync cause one? 


ok now back to postfix and spam. Jorn, how would spamassassin help my 
problem, the spam is not getting to any of my accounts, since postfix is 
bouncing the mail back before it ever reaches a user mailbox. although 
maybe spam assassin does something i was unaware of. if it does have some 
sort of black list feature or something else that can help me great. 
please enlighten me.

and Bill, ill look into the greylist you talked about, it looks really 
helpful. my only question is will it significantly reduce the network 
traffic? right now my problem is not accepting the mail, its just that my 
personal server on dsl doesnt have the bandwidth to bounce all of these 
messages back. also does the grey listing have any advandtages to using a 
black list provided by spamhaus.org or something similar.

thank you both for your help
anthony philipp


On Wed, 2 Jun 2004, Jorn Argelo wrote:

> Anthony Philipp wrote:
> 
> >hello,
> >i am running a personal server and after nearly 26 days of uptime without any major problems last night at around ten pm. central time there was what i assume is a kernel panic. im running 5.2 RELEASE on a pentium 3 877 mhz. 
> >heres the output: (hopefully this is accurate i had to write this out by hand, if someone knows how to get the orignal output that would be nice)
> >
> >Fatal trap 12: page fault while in kernel mode
> >cpuid =0; apic id = 00
> >fault virtual address = 0x38
> >fault code = supervisor read, page not present
> >instruction pointer	= 0x8:0xc06bc08a
> >stack pointer		= 0x10:0xcb57098c
> >frame pointer		= 0x10:0xcb570a68
> >code segment		= base 0x0, limit 0xfffff, type 0x1b
> >			= DPL 0, pres1, dres321, gran1
> >processor eflags	= interrupt enabled, resume, 10pl=0
> >current process		= 27 (swi1:net)
> >trap number		= 12
> >panic page fault
> >cpuid=0;
> >
> >syncing disks, buffers remaining... 1787 1787
> >
> >and then it was crashed hard. i couldnt switch virtual terminals, i couldnt shut down, all i could do is press the power button. so thats my first problem, what happened? was it a kernel panic? or was it something different, and whatever it was, what caused it? 
> >  
> >
> Yup, it was a kernel panic. Some process caused your panic, which is in 
> this case swi1:net. Any idea what it is?
> 
> >
> >ok now onto the next problem. i was running my own mail server until recently as well, but i was getting beaten into the ground by spam. my /var/spool/postfix directory was 1.8 gigabytes. it was all spam i was bouncing from dynamic ips. none of it was getting to my accounts, but my server was getting clogged, since im running this off of adsl. im just wondering what i can do to stop this. the legitimate mail was not a problem at all. also this only started being a problem after a few weeks. as a temporary solution ive stopped postfix and deleted my entire spooled mail directory. however i would like a more permenant solution to these spammers attacking my server. right now postfix is bouncing all emails where the ips dont match the domain name. 
> >  
> >
> Consider installing spamassassin?
> 
> Cheers,
> 
> Jorn
> 

-- 



More information about the freebsd-questions mailing list