ax25/netrom/crashes?

Hi!
I am (still) struggling with crashes.

I now suspect, even more, netrom has to do with the crashes, latest hint is this:
From /var/messages I can se one user logging in to dxspider who probably has come over so many 1200 bit/s nodes that it cannot pass the traffic fast enough, but a (netrom transport) queue is built up. After 1,5 hour of connection, the node is crashing.

- Maybe due to filled up queue (not liekely)?
- Maybe due to a disconnect from the remote end with a large queue?

Below is the tail -f of the log file as seen over 1200 bit/s, including the disconnect after the crash:

Aug  4 19:18:24 router-sk7hw ax25d[2733]: NET/ROM SK6HD-0@SK7IJ-5 (netrom) node
Aug  4 19:18:24 router-sk7hw node[2733]: SK6HD @ SK7IJ-5 logged in
Aug  4 19:19:08 router-sk7hw ax25d[2734]: NET/ROM SK6HD-0@SK7HW-5 (netsp) client
Aug  4 19:19:08 router-sk7hw node[2733]: Connected to DXSVXO:SK7HW-7

*** Cleared
Mon Aug 4 20:59:17 CEST 2003
[sm7gvf@pc2 sm7gvf]$

kernel 2.4.22-pre10
HR6.1

Guessing here (the system is down):
libax25-devel-0.0.7
libax25-0.0.7
ax25-tools-0.0.6
ax25-apps-0.0.4

running dxnet, fbb, dxspider, node0.3.0

I am using the script below as startup, I don't have all the other files at hand here.

Could it be netrom causing this?

What more information could I supply?

73
Kjell, sm7gvf

#!/bin/sh
#
# ax25init This inits the adx25 interfaces and services
#
# chkconfig: - 94 03
# description: ax25 init script

# Source function library.
. /etc/rc.d/init.d/functions

# Source networking configuration.
. /etc/sysconfig/network

# Check that networking is up.
[ ${NETWORKING} = "no" ] && exit 0

#DAEMON=ax25d

# See how we were called.
case "$1" in
start)

        # Start daemons.
        echo -n "Starting ax25 interfaces: "
	
	mkiss -s 9600 /dev/ttyS0 /dev/ptyq0
	kissattach /dev/ttyq0 emma 44.140.249.10
	ifconfig  ax0 netmask 255.255.255.255
	kissparms -p emma -t 200 -r 255 -s 10

	mkiss -c -p 2 -s 9600 /dev/ttyS1 /dev/ptyq1
	kissattach /dev/ttyq1 gno 44.140.249.10
	ifconfig ax1 netmask 255.255.255.255
	kissparms -p gno -f n -t 100 -r 255 -s 10

	sccinit
	
	ifconfig scc0 44.140.249.10 netmask 255.255.255.255 up
	axparms -setcall scc0 sk7hw-10
	ifconfig scc1 44.140.249.10 netmask 255.255.255.255 up
	axparms -setcall scc1 sk7hw-11
	ifconfig scc2 44.140.249.10 netmask 255.255.255.0 up
	axparms -setcall scc2 sk7hw-9
	ifconfig scc3 44.140.249.10 netmask 255.255.255.255 up
	axparms -setcall scc3 sk7hw-13
	
	sccparam scc0 wait 10
	ifconfig scc0 txqueuelen 20
	sccparam scc1 wait 10
	ifconfig scc1 txqueuelen 20
	sccparam scc2 wait 10
	ifconfig scc2 txqueuelen 20
	sccparam scc2 maxkeyup 8
	sccparam scc3 wait 10
	ifconfig scc3 txqueuelen 20
	
	nrattach -i 44.140.249.10 netrom
	ifconfig nr0 netmask 255.255.255.255
	nrattach -i 44.140.249.10 netfbb
	ifconfig nr1 netmask 255.255.255.255
	nrattach -i 44.140.249.10 netdx
	ifconfig nr2 netmask 255.255.255.255
	nrattach -i 44.140.249.10 netsp
	ifconfig nr3 netmask 255.255.255.255
	
	echo 24000	> /proc/sys/net/netrom/transport_timeout
	echo 2		> /proc/sys/net/netrom/transport_maximum_tries
	echo 2000	> /proc/sys/net/netrom/transport_acknowledge_delay

#emma

	echo 2     	> /proc/sys/net/ax25/ax0/standard_window_size
	echo 0		> /proc/sys/net/ax25/ax0/backoff_type
	echo 800	> /proc/sys/net/ax25/ax0/t1_timeout 
 	echo 300	> /proc/sys/net/ax25/ax0/t2_timeout
	echo 20	 	> /proc/sys/net/ax25/ax0/maximum_retry_count

#gnosjo

	echo 2     	> /proc/sys/net/ax25/ax1/standard_window_size
	echo 0		> /proc/sys/net/ax25/ax1/backoff_type
	echo 600	> /proc/sys/net/ax25/ax1/t1_timeout
	echo 200	> /proc/sys/net/ax25/ax1/t2_timeout
        echo 20         > /proc/sys/net/ax25/ax1/maximum_retry_count

#ryda

	echo 2    	> /proc/sys/net/ax25/scc0/standard_window_size
	echo 0		> /proc/sys/net/ax25/scc0/backoff_type
        echo 900       > /proc/sys/net/ax25/scc0/t1_timeout
	echo 300	> /proc/sys/net/ax25/scc0/t2_timeout
        echo 20         > /proc/sys/net/ax25/scc0/maximum_retry_count

#Ekekull

	echo 2     	> /proc/sys/net/ax25/scc1/standard_window_size
	echo 0		> /proc/sys/net/ax25/scc1/backoff_type
        echo 400	> /proc/sys/net/ax25/scc1/t1_timeout
	echo 100	> /proc/sys/net/ax25/scc1/t2_timeout
        echo 20         > /proc/sys/net/ax25/scc1/maximum_retry_count

#Lan

	echo 2     	> /proc/sys/net/ax25/scc2/standard_window_size
	echo 0		> /proc/sys/net/ax25/scc2/backoff_type
        echo 1300       > /proc/sys/net/ax25/scc2/t1_timeout
	echo 300	> /proc/sys/net/ax25/scc2/t2_timeout
        echo 20         > /proc/sys/net/ax25/scc2/maximum_retry_count
	echo 0		> /proc/sys/net/ax25/scc2/ip_default_mode
	echo 400	> /proc/sys/net/ipv4/neigh/scc2/retrans_time 

#Asarum

	echo 2     	> /proc/sys/net/ax25/scc3/standard_window_size
	echo 0		> /proc/sys/net/ax25/scc3/backoff_type
        echo 1000       > /proc/sys/net/ax25/scc3/t1_timeout    
	echo 300	> /proc/sys/net/ax25/scc3/t2_timeout
        echo 20         > /proc/sys/net/ax25/scc3/maximum_retry_count

	mheardd -n 100
	ax25d -l
	netromd -i -t 30 -l -d

# rxecho

	. /var/ax25/nodesave.data
	. /etc/ax25/routes.conf &
	axdigi &
				
        RETVAL=$?
        echo
        ;;

stop)
# Stop daemons.
echo -n "Shutting down $DAEMON: " /usr/sbin/nodesave /var/ax25/nodesave.data killall mkiss
killall kissattach
killall mheardd
killall ax25d
killall rxecho
killall netromd
ifconfig scc0 down
ifconfig scc1 down
ifconfig scc2 down
ifconfig scc3 down
ifconfig nr0 down
ifconfig nr1 down
ifconfig nr2 down
ifconfig nr3 down
ifconfig ax0 down
ifconfig ax1 down

RETVAL=$?
echo
[ $RETVAL -eq 0 ]
;;

status)

        status $DAEMON
        RETVAL=$?
        ;;

restart|reload)

        $0 stop
        $0 start
        RETVAL=$?
        ;;

*)
echo "Usage: $DAEMON {start|stop|restart|status}" exit 1
esac

exit $RETVAL

Hi

When I was using RedHat 7.2 I was having crashes once or twice a day. The majority of them were after a NetRom disconnect. I have now reverted to Red Hat 6.2 with Kernel 2.2.24 and my node (GB7WDX) has been up for 112 days.

73 de John, G7OPJ

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Kjell Jarl
Sent: 04 August 2003 21:10
To: Linux Hams
Subject: ax25/netrom/crashes?

Hi!
I am (still) struggling with crashes.

I now suspect, even more, netrom has to do with the crashes, latest hint is this:
From /var/messages I can se one user logging in to dxspider who
probably has come over so many 1200 bit/s nodes that it cannot pass the traffic fast enough, but a (netrom transport) queue is built up. After 1,5 hour of connection, the node is crashing.

- Maybe due to filled up queue (not liekely)?
- Maybe due to a disconnect from the remote end with a large queue?

Below is the tail -f of the log file as seen over 1200 bit/s, including the disconnect after the crash:

Aug  4 19:18:24 router-sk7hw ax25d[2733]: NET/ROM SK6HD-0@SK7IJ-5 
(netrom) node
Aug  4 19:18:24 router-sk7hw node[2733]: SK6HD @ SK7IJ-5 logged in
Aug  4 19:19:08 router-sk7hw ax25d[2734]: NET/ROM SK6HD-0@SK7HW-5 
(netsp) client
Aug  4 19:19:08 router-sk7hw node[2733]: Connected to DXSVXO:SK7HW-7

*** Cleared
Mon Aug 4 20:59:17 CEST 2003
[sm7gvf@pc2 sm7gvf]$

kernel 2.4.22-pre10
HR6.1

Guessing here (the system is down):
libax25-devel-0.0.7
libax25-0.0.7
ax25-tools-0.0.6
ax25-apps-0.0.4

running dxnet, fbb, dxspider, node0.3.0

I am using the script below as startup, I don't have all the other files at hand here.

Could it be netrom causing this?

What more information could I supply?

73
Kjell, sm7gvf

Same here, RedHat 6.2 was perfect, no problems whatsoever. RedHat 7.3 on the other hand crashes, sometimes, on netrom and ax25 disconnects. Sometimes several times a day, sometimes it is up for almost a month.

It crashes in the interrupt handler.

I had some log and listen excerpts but I can not find them now...

So something in the 2.4-kernels is creating problems... Hopefully 2.6 will work better...

73 de Lars, sm6rpz

Lars E. Pettersson, SM6RPZ | Gothenburg, SWEDEN
[email protected] | http://www.sm6rpz.se/

Using a standard 2.4.21 kernel, there are no problems with AX25.

But the system sometimes crashes the moment I disconnect from a netrom node that I have previously connected to.

I tried the 9A4GL patch that was posted recently, but it wouldn't apply against the 2.4.21 kernel. Has anyone successfully done this?

The only other problem I have is, that the parallel port Baycom driver ('picpar') cannot be removed without a kernel panic. So, when I need the parallel port for something else, it requires the "Microsoft Soulution" (a reboot).

Wilbert, ZL2BSJ

<snip>

<snip>