[Lightyear]

Normally, I am very respectful of code I don’t quite understand. In this case, I was forced to investigate because for me Zeroshell is mission-critical and I could not wait for a patch.

So, you have been warned, continue reading at your own risk 😐

Perhaps the best bet would be to add

chmod 644 /root/kerbynet.cgi/scripts/autoupdater

to the pre boot script.
Of course, this would stop all autoupdates, but I really can’t say what unwanted effects it would also bring.
In any case, better try it with a disposable backup 😉

[Lightyear]

In the meanwhile you can ssh to zeroshell and go to the shell prompt (option S)
Then: cd /var/register/system/AutoUpdate/pkgs/
There it should be directory 23321 with the troublesome patch. Rename it to anything lesser than 23320:

mv 23321 23319

Sadly I don’t remember if I erased 23320; I believe I didn’t, but if necessary, you could do

rm -rdf 23320

Then exit the shell prompt and log in via the web interface. Now you can happily reinstall 23320.

Just in case 23321 tries to make a comeback, let’s make a file with the same name.
Go back to the shell prompt and cd /var/register/system/AutoUpdate/pkgs/
Then,

touch 23321

And, to be on the safe side, make it unwritable at least as long as it is not chmodded:

chmod 444 23321

Of course, instead of renaming you could erase 23321. I believe it should work, but I really don’t know.

Hope this helps, it worked for me.

[Lightyear]

Went back to 3.2.1 and it gave the same error.
Further investigation determined that the pendrive with the configuration has some errors and the file /Database/var/register/system/ssl/ForceHTTPS was missing.
Very sorry for the mistake 😳 . Perhaps it may be of help to someone, though.
Anyway, ZS rocks! 😀

[Lightyear]

Same behavior here. Going back to 3.2.1
Thank you erick06fr for your thorough work! Saved me a lot of time!

[Lightyear]

Restarting did the trick. When Zeroshell came back, all was fine.

[Lightyear]

No need to bother Ivan; here follows the modified script 8)

#!/bin/sh

# Script made by Ivan Baldo to permit multiple dyndns accounts in ZeroShell



Host_PPP0="http_address_of_dyndns_account_1"

Host_PPP1="http_address_of_dyndns_account_2"



actualiza ()

{

Interfaz="$1"

Host="$2"

Usuario="$3"

Contrasena="$4"

IP=$(ifconfig $Interfaz 2>/dev/null | awk 'NR==2 {print $2}' | grep '.*..*..*..*' | cut -d: -f2)

touch /tmp/dyndns.ipviej.$Interfaz

IPViej="$(cat /tmp/dyndns.ipviej.$Interfaz)"

Actualizar="no"

if [ "$IP" != "" ]; then

if [ "$IP" != "$IPViej" ]; then

Actualizar="si"

fi

fi

if [ "$Actualizar" = "no" ]; then

return

fi

if [ -e /tmp/dyndns.deshabilitado.$Interfaz ]; then

return

fi

logger -t DDNS "Actualizando $Host a $IP de interfaz $Interfaz."

if wget "--bind-address=$IP" -t 3 -w 20 -o /tmp/dyndns.error.$Interfaz -O /tmp/dyndns.resultado.$Interfaz 

"http://$Usuario:$Contrasena@members.dyndns.org/nic/update?system=dyndns&hostname=$Host&mx=$Host&myip=$IP"

then

logger -t DDNS "==> $(cat /tmp/dyndns.resultado.$Interfaz)"

if grep -q "!yours|notfqdn|abuse|nohost|badagent|badauth|badsys|dnserr|numhost|911|!donator" 

/tmp/dyndns.resultado.$Interfaz

then

logger -t DDNS "==> Hubo un error reportado por DynDNS, luego de solucionar el problema borre el archivo /tmp/dyndns.deshabilitado.$Interfaz!"

touch /tmp/dyndns.deshabilitado.$Interfaz

else

echo "$IP" >/tmp/dyndns.ipviej.$Interfaz

fi

else

logger -t DDNS "==> Error al conectarse con DynDNS:"

grep -v " => " /tmp/dyndns.error.$Interfaz | logger -t DDNS

fi

}



while true; do

actualiza ppp0 $Host_PPP0 username_1 password_1

actualiza ppp1 $Host_PPP1 username_2 password_2

sleep 60

done 

Of course, to provide for another DDNS service, it would be neccesary to know these other service quirks and tricks.
And thank you for the tip, ppalias! I do hope you have had a very good 2010 beginning.

[Lightyear]

I currently have two dynamic IP connections, one ADSL and one 3G.
Ivan Baldo, fine programmer and friend extraordinaire, made a script so ZeroShell updates both dynamic IPs.
He called it ddns2.nuestro (we are Spanish speakers) and placed it in the /Database directory.
Then he added the following command to the ZeroShell preboot script:

cp -v /Database/ddns2.nuestro /root/kerbynet.cgi/scripts/ddns2

so ddns2.nuestro is what ZeroShell wil run when updating dynamic IPs.
It has worked without a glitch for more than 3 months now, and I think it is very easy to modify for more than 2 dyndns accounts, or for another DDNS service.
Here it goes:

#!/bin/sh

# Script made by Ivan Baldo to permit multiple dyndns IP updating in ZeroShell

Usuario=”your_dyndns_user_name”
Contrasena=”your_dyndns_password”
Host_PPP0=”http_address_of_dyndns_account_1″
Host_PPP1=”http_address_of_dyndns_account_2″

actualiza ()
{
Interfaz=”$1″
Host=”$2″
IP=$(ifconfig $Interfaz 2>/dev/null | awk ‘NR==2 {print $2}’ | grep ‘.*..*..*..*’ | cut -d: -f2)
touch /tmp/dyndns.ipviej.$Interfaz
IPViej=”$(cat /tmp/dyndns.ipviej.$Interfaz)”
Actualizar=”no”
if [ “$IP” != “” ]; then
if [ “$IP” != “$IPViej” ]; then
Actualizar=”si”
fi
fi
if [ “$Actualizar” = “no” ]; then
return
fi
if [ -e /tmp/dyndns.deshabilitado.$Interfaz ]; then
return
fi
logger -t DDNS “Actualizando $Host a $IP de interfaz $Interfaz.”
if wget “–bind-address=$IP” -t 3 -w 20 -o /tmp/dyndns.error.$Interfaz -O /tmp/dyndns.resultado.$Interfaz
“http://$Usuario:$Contrasena@members.dyndns.org/nic/update?system=dyndns&hostname=$Host&mx=$Host&myip=$IP”
then
logger -t DDNS “==> $(cat /tmp/dyndns.resultado.$Interfaz)”
if grep -q “!yours|notfqdn|abuse|nohost|badagent|badauth|badsys|dnserr|numhost|911|!donator”
/tmp/dyndns.resultado.$Interfaz
then
logger -t DDNS “==> Hubo un error reportado por DynDNS, luego de solucionar el problema borre el archivo /tmp/dyndns.deshabilitado.$Interfaz!”
touch /tmp/dyndns.deshabilitado.$Interfaz
else
echo “$IP” >/tmp/dyndns.ipviej.$Interfaz
fi
else
logger -t DDNS “==> Error al conectarse con DynDNS:”
grep -v ” => ” /tmp/dyndns.error.$Interfaz | logger -t DDNS
fi
}

while true; do
actualiza ppp0 $Host_PPP0
actualiza ppp1 $Host_PPP1
sleep 60
done

Unfortunately, the forum editor does not allow script formatting. I hope it is readable all the same.
I believe also it’s easy to translate the Spanish parts if need be, but in any case I will be following this thread so feel free to ask for translation.

I hope I didn’t break any forum rule copying the script here, but I feel it really can be of help to someone, and I do not have any place to hang it for downloading 🙁 .
ZeroShell rules! 😀

[Lightyear]

What works for me is ssh-ing to ZeroShell and then go to the ZeroShell system prompt. From there you can scp to/from anywhere in your network that has password authentication.
And, if you want key authentication, just scp the keys to ZeroShell.
This is what I use to backup ZeroShell to a local server. And what I used to bring Joe to ZeroShell.
Admittedly, mine is a Linux-only network, but from what I have heard from the Mac, it should be a piece o’cake.

Hth.

[Lightyear]

Never tried FTP. What i’m doing is using scp to back up to a local machine.

Hth.

[Lightyear]

Thank you for your reply, I wasn’t expecting one so fast in Christmas 🙂
Yes, I have a virtual server for each connection, and I route all P2P via 3g modem (ppp1). That is working fine.
What leaves me at a loss is that even if I put a balancing rule to send all the traffic (no protocol filter, no port filter, only “from” address filter) that comes from the machine running MLDonkey via ppp1, MLDonkey keeps using the wrong IP address (the one given by the ISP to ppp0, adsl connection) to make a port test.
This happens roughly 60% of the time (the weight of the adsl gateway).
Anyway, I have come to a solution of sorts setting MLDonkey to discard connections with lowID. Crude but effective.
I would have liked to know what’s going on, though. 😕
But it’s a very minor point. ZeroShell rules! 😀

Merry Christmas to you, ppalias!

[Lightyear]

Try

http://www.zeroshell.net/eng/forum/viewtopic.php?t=1645

Hth.

[Lightyear]

My problem is slightly different. I’m using MLDonkey and via Virtual servers/Net balancing it’s working fairly well.
The only thing to improve is that there’s no way to know which address will ZeroShell return when MLDonkey makes a port test. So, it’s fairly common for MLDonkey to test the wrong address and get lowID.
The connections are dynamic IP, so there is no way to know the addresses beforehand.
I believe it is a different problem from that of the previous poster, but I’m not sure.
Thank you in advance.

[Author]

Posts