Hello I am running ntop with a hand compiled 2.4.23 kernel.  I have sucessfully applied PF_RING, and RTIRQ patches, to the kernel itself.
 
Linux nflowbox 2.4.23 #2 SMP Sat Dec 4 18:33:51 PST 2004 i686 i686 i386 GNU/Linux
 
Ntop runs beautifully except after awhile, it will crash.
 
It is sitting on three nic's, One for direct access to the box itself, one Nic to support the netflow network, and one nic to dump data out of a portmirrored uplink port on a switch.
 
It should be handling about 5~33Mbits of outbound and 3~5 inbound traffic on the port mirrored switchport.  And about roughly 7000~9000 packets a second with netflow.
 
The problem I am getting is when I am viewing ntop, and it ends up timing out, which ends up crashing the whole machine itself. See the /var/log/messages snippet at the bottom.
 
Dec  5 12:41:43 nflow ntop[4243]:   **ERROR** http generation failed, alarm() tripped. Please report this to ntop-dev list!
Dec  5 12:41:46 nflow ntop[4244]:   **ERROR** http generation failed, alarm() tripped. Please report this to ntop-dev list!
Dec  5 12:50:50 nflow ntop[4248]:   **ERROR** http generation failed, alarm() tripped. Please report this to ntop-dev list!
Dec  5 13:20:30 nflow kernel: __alloc_pages: 0-order allocation failed (gfp=0x1d2/0)
Dec  5 14:52:49 nflow last message repeated 4 times
Dec  5 14:55:39 nflow kernel: __alloc_pages: 0-order allocation failed (gfp=0x1d2/0)
Dec  5 14:55:39 nflow kernel: VM: killing process ntop
Dec  5 14:55:39 nflow kernel: __alloc_pages: 0-order allocation failed (gfp=0x1d2/0)
Dec  5 14:55:40 nflow kernel: __alloc_pages: 0-order allocation failed (gfp=0x1d2/0)
Dec  5 14:57:10 nflow ntop[4257]:   **ERROR** http generation failed, alarm() tripped. Please report this to ntop-dev list!
Dec  5 19:24:13 nflow sshd(pam_unix)[4278]: session opened for user root by (uid=0)
Libpcap-ring has been compiled,  with tcpdump and ntop libraries tied in. Version of libpcap 0.8.1 tcpdump is 3.8.1 and ntop 3.0 tgz from sourceforge.
 
I dont know what could be causing it, so im up for any suggestions.  Im thinking RAM but i have 1GB worth, and 1GB of swap on another drive.  So, im going to tail the ram output for awhile to see if it does use up all of it. in which, could be a mem leak?  Or maybe im pushing this too far.

[EMAIL PROTECTED] log]# cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 15
model           : 2
model name      : Intel(R) Celeron(R) CPU 2.60GHz
stepping        : 9
cpu MHz         : 2600.474
cache size      : 128 KB
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 2
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe cid
bogomips        : 5190.45
 
[EMAIL PROTECTED] log]# free -m
             total       used       free     shared    buffers     cached
Mem:          1000         27        973          0          5          8
-/+ buffers/cache:         13        987
Swap:         1000          4        995
[EMAIL PROTECTED] log]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/hda5             4.9G  339M  4.3G   8% /
none                  4.9G  339M  4.3G   8% /sys
/dev/hda1             194M   12M  173M   7% /boot
none                  501M     0  501M   0% /dev/shm
/dev/hda7              54G  1.5G   50G   3% /home
/dev/hda2             7.7G  2.9G  4.5G  40% /usr
/dev/hda3             7.7G  376M  7.0G   6% /var
/dev/hdc1             113G   42G   65G  40% /usr/local/var
[EMAIL PROTECTED] log]#
 

======================
Israel Lopez
OCHosting Inc
Systems Administrator
http://www.ochosting.com
_______________________________________________
Ntop-dev mailing list
[EMAIL PROTECTED]
http://listgateway.unipi.it/mailman/listinfo/ntop-dev

Reply via email to