Re: remote telnet is blocked when internal ping is ongoing (re-submit)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hello,

	The problem is most likely not related to iptables and is related to the network card. What type of network card are
you using ?? Do you know the buffer sizes ??

If this nic is a server class network card the send and receive buffers should almost be equal. You will find with
desktop network cards that the receive buffer is twice or three times greater then the send buffer. You are most likely
"flooding" the network card with your large ping packets at that rate ... causing it to not be able to accept the telnet
connection or at least slow it down.

Michael.



On Tue, 29 Jun 2004 10:34:36 -0400
"Linda Zhu" <lindazhu@xxxxxxxxxx> wrote:

> Hi,
> 
> This is Linda from UTStarcom. I have problems with iptable configuration.
> 
> The problem happens on one card. The card has two ethernet interfaces:
> One is 172.16.xx.xx, and another is 192.168.xx.xx. The 172.16.xx.xx
> interface
> is used for remotely acceccing the card. The 192.168.xx.xx interface
> receives
> and transmit traffic.
> 
> Part of our iptable configuration is attached below.
> 
> In normal case, user can remotely telnet into this card with no problme.
> The problem occurs when we run ping from 196.168.xx.xx interface at a fast
> rate.
> 
> e.g. ping -i 0.01 -s 1400 192.168.100.2, then the remote telnet to
> 172.168.xx.xx
> interface is blocked.
> 
> In the iptable configuration, we have one rule:
> 
> $IPTABLES -A allowed_tcp -p TCP --syn -m limit --limit 1/second -j ACCEPT
> 
> If we change the limit from 1/second to 20/second, then with one ping
> session,
> the telnet will be successful. After that, if we run two ping sessions:
> e.g.
> ping -i 0.01 -s 1400 192.168.100.2
> ping -i 0.01 -s 1400 192.168.100.3
> 
> then telnet is blocked again.
> 
> Then if we change the limit from 20/second to 100/second, remote telnet
> success.
> With the 100/second, even 4 ping sessions ongoing at same time, the remote
> telnet is OK.
> 
> Basically, we don't setup any firewall for 192.168.xx.xx interface. My
> question
> is why the ping on 192.168.xx.xx interface will affect the remote telnet.
> The two
> interface are in different IP groups and on different hardware. Please
> advice.
> 
> Thank you,
> Linda
> 
> ================== part of our iptable configuration =======================
> 
> ###
> # Create separate chains for TCP,UDP,ICMP to traverse
> #
> #
> 
> $IPTABLES -N allowed_tcp
> $IPTABLES -N tcp_packets
> $IPTABLES -N udp_packets
> $IPTABLES -N icmp_packets
> #$IPTABLES -N allowed_ping
> 
> #
> # Allowed chain
> #
> 
> $IPTABLES -A allowed_tcp -p TCP --syn -m limit --limit 1/second -j ACCEPT
> $IPTABLES -A allowed_tcp -p TCP -m state --state ESTABLISHED,RELATED -j
> ACCEPT
> $IPTABLES -A allowed_tcp -p TCP -j DROP
> 
> #
> # TCP rules
> #
> 
> ### Telnet ###
> $IPTABLES -A tcp_packets -p TCP -s 0/0 --dport 23 -j allowed_tcp
> 
> 
> #
> # ICMP rules
> #
> 
> $IPTABLES -A icmp_packets -p ICMP -s 0/0 -m limit --limit
> 3/second --icmp-type 8
>  -j ACCEPT
> $IPTABLES -A icmp_packets -p ICMP -s 0/0 --icmp-type pong -j ACCEPT
> $IPTABLES -A icmp_packets -p ICMP -s 0/0 --icmp-type port-unreachable -j
> ACCEPT
> $IPTABLES -A icmp_packets -p ICMP -s 0/0 --icmp-type 11 -j ACCEPT
> 
> 
> 
> 
> 
> 


-- 
Michael Gale
Network Administrator
Utilitran Corporation


[Index of Archives]     [Linux Netfilter Development]     [Linux Kernel Networking Development]     [Netem]     [Berkeley Packet Filter]     [Linux Kernel Development]     [Advanced Routing & Traffice Control]     [Bugtraq]

  Powered by Linux