Cron - it's not working

eXaLtY · 30. Juli 2012, 18:59

Hi Guys,

I have the new 3.3.2 image installed in my dm800se, I have installed the cron via telnet and added the following:
(all 755)
/usr/script/reboot.sh

#!/bin/sh
wget -q -O - root:password@192.168.12.3/web/powerstate?newstate=2

I've tried this as well :

/sbin/init 5
/bin/sleep 5
/sbin/reboot

/cron/tabs/root

0 4 * * * /usr/script/reboot.sh

when I do crontab -l

the cron is listed.

But doesn't reboot!
I've also tried to start the cron manually but still doesn't work.

Any idea?

Fred Bogus Trumper · 30. Juli 2012, 21:15

Does your script work if started manually? Is the file-attribute of your script 755? Did you use a linux/unix editor to create?

There was/is a bug in the DMM Web-Interface too. Some commands like reboot, restart didn't work when executed from shell or by script. I'm not shure that this bug is already fixed in Newnigma2 3.3.2 or if the DMM bug-fix for the Web-Interface was issued after NN2 v3.3.2 release.

Try to execute the command twice in your script:

Shell-Script

#!/bin/sh
wget -q -O - http://root:password@192.168.12.3/web/powerstate?newstate=2
wget -q -O - http://root:password@192.168.12.3/web/powerstate?newstate=2

this "fix" worked in one of my own shellscripts. In current images based on OE2.0 this bug is definitely removed

eXaLtY · 31. Juli 2012, 08:33

Hi,

Thanks for replying.

The chmod are in deed 755 and I've used the DCC editor.

Is not the 3.3.2 OE2.0 based?
How can I see that?
And where are the OE2.0 images from newnigma?

I will try today the command twice.

eXaLtY · 1. August 2012, 16:01

It doesn't work.

Fred Bogus Trumper · 1. August 2012, 18:44

do you use the plugin "cron" from the online feed? the dmm busybox-cron in OE1.6 is buggy too:

try this to test cron

opkg update
opkg install cron
crontab -e

Quellcode

*/1 * * * * /bin/echo "cron test ok - $(date)" >> /tmp/crontest.log

save the new crontab and restart cron

/etc/init.d/cron stop
/etc/init.d/cron start

the log should look like this after a few minutes:

Quellcode

root@dm800se:~# cat /tmp/crontest.log
cron test ok - Wed Aug 1 18:41:00 CEST 2012
cron test ok - Wed Aug 1 18:42:00 CEST 2012
cron test ok - Wed Aug 1 18:43:00 CEST 2012
root@dm800se:~#

and you can be shure that cron works correctly, next step is your script

eXaLtY · 1. August 2012, 19:46

Thanks for the reply, I've followed your instructions and still no log file in TMP folder.

in /cron/log I HAVE:

root (08/01-19:38:04-27680) BEGIN EDIT (root)
CRON (08/01-19:39:47-27962) STARTUP (fork ok)

Fred Bogus Trumper · 2. August 2012, 00:20

cron does his job in Newnigma2 v3.3.2, your log tells you

Quellcode

CRON (08/01-19:39:47-27962) STARTUP (fork ok)

try this to check cron is running

root@dm800se:~# ps | grep -v grep | grep cron
2148 root 2044 S /usr/sbin/cron
root@dm800se:~#

my log in Newnigma2 v.3.3.2

Quellcode

root@dm800se:~# cat /cron/log
CRON (08/02-00:07:14-2148) STARTUP (fork ok)
root (08/02-00:08:00-2279) CMD (/usr/script/cronmessage.sh)
root (08/02-00:08:00-2281) CMD (/bin/echo "cron test ok - $(date)" >> /tmp/crontest.log)
root (08/02-00:09:00-2480) CMD (/usr/script/cronmessage.sh)
root (08/02-00:09:00-2482) CMD (/bin/echo "cron test ok - $(date)" >> /tmp/crontest.log)
root (08/02-00:10:00-2672) CMD (/usr/script/cronmessage.sh)
root (08/02-00:10:00-2674) CMD (/bin/echo "cron test ok - $(date)" >> /tmp/crontest.log)
root (08/02-00:11:00-2859) CMD (/usr/script/cronmessage.sh)
root (08/02-00:11:00-2860) CMD (/bin/echo "cron test ok - $(date)" >> /tmp/crontest.log)
root (08/02-00:12:00-3064) CMD (/usr/script/cronmessage.sh)
root (08/02-00:12:00-3065) CMD (/bin/echo "cron test ok - $(date)" >> /tmp/crontest.log)
root@dm800se:~#

Alles anzeigen

there must be a mistake in your script(s). Did you install v3.3.2 in Flasch or do u use tools like Barry Allen, Dumbo, Rambo or things like this to boot from USB or to expand your Flash?

eXaLtY · 3. August 2012, 09:18

Thanks for your reply!

I've checked and cron log as this info:

CRON (08/02-06:55:45-649) STARTUP (fork ok)
root (08/02-19:31:06-28757) BEGIN EDIT (root)
root (08/02-19:37:46-28757) REPLACE (root)
root (08/02-19:37:46-28757) END EDIT (root)
root (08/02-19:38:00-649) RELOAD (tabs/root)
CRON (08/02-19:39:42-30347) STARTUP (fork ok)
root (08/03-04:00:00-16682) CMD (/usr/script/reboot.sh)
CRON (08/03-05:50:37-651) STARTUP (fork ok)

So it means that cron is working fine....

I've installed in the flash directly.

So maybe something with the script?

#!/bin/sh
wget -q -O - root:password@192.168.23.13/web/powerstate?newstate=2
wget -q -O - root:password@192.168.23.13/web/powerstate?newstate=2

I've made a duplicate like you told me but still doesn't work.

Fred Bogus Trumper · 3. August 2012, 10:56

I testetd the reboot executed by cron in NN2 v3.3.2 and it works - the Web-IF bug is already fixed and only one command is necessary

try this on commandline to create the (new) script
copy & paste - 127.0.0.1 as localhost instead of IP of the Box works too

Quellcode

echo "#!/bin/sh" > /usr/script/reboot.sh
echo "wget -q -O- http://root:password@127.0.0.1/web/powerstate?newstate=2" >> /usr/script/reboot.sh
chmod 755 /usr/script/reboot.sh

eXaLtY · 4. August 2012, 11:44

Mate,

It's Working!

CRON (08/04-00:34:33-651) STARTUP (fork ok)
root (08/04-00:54:26-3313) LIST (root)
CRON (08/04-01:06:38-657) STARTUP (fork ok)
CRON (08/04-01:09:29-652) STARTUP (fork ok)
root (08/04-04:00:00-23829) CMD (/usr/script/reboot.sh)
CRON (08/04-04:00:51-653) STARTUP (fork ok)
CRON (08/04-06:09:32-650) STARTUP (fork ok)

I wake up and I saw up time since 4 AM!

Many thanks!!!!

I only saw this WARNING: invalid client tcp_conn_close()
2012/08/04 06:03:13 2B702528 c cards disconnected from 127.0.0.1

In Oscam after the reboot couldn't pick up the user from CCcam.

Do you think that might be related??
Is there any other way to reboot without using the web interface command?

Thanks dude, really nice of you!

Fred Bogus Trumper · 4. August 2012, 15:47

Original von eXaLtY
Is there any other way to reboot without using the web interface command?

reboot
or better, I think
shutdown -r now

and u don't need a script for this, just use a crontab

0 4 * * * /sbin/shutdown -r now

if u want the reboot in cause of problem with your emu: there is an option in Newnigma2 Services for daily emu restart

or executed wiht cron

0 4 * * * /usr/bin/camdctrl restart

this restarts the running emu

eXaLtY · 4. August 2012, 18:54

Dude,

I don't know if someone already told you? But You Rock my Friend!!!!

Does it work as well in OE2.0 images?

Do you have a script for checking if oscam is not running then restart camd?

EverythingZen · 6. August 2012, 00:14

oscamcheck.sh

Shell-Script

#!/bin/sh
if ps -a |grep -v grep |grep -c oscam_1.20 >/dev/null
then
echo "OScam ok"
else
/usr/bin/oscam_1.20 -b
fi

cron

Quellcode

*/2 * * * * /usr/script/oscamcheck.sh

you may have to change oscam version in the script, in this example v. 1.20

i guess a well configured oscam doesn't need something like this...

eXaLtY · 6. August 2012, 12:39

Originally posted by EverythingZen
oscamcheck.sh

Shell-Script

#!/bin/sh

if ps -a |grep -v grep |grep -c oscam_1.20 >/dev/null

then

echo "OScam ok"

else

/usr/bin/oscam_1.20 -b

fi

cron

Quellcode

*/2 * * * * /usr/script/oscamcheck.sh

you may have to change oscam version in the script, in this example v. 1.20

i guess a well configured oscam doesn't need something like this...

Would this also work?

#!/bin/sh
if ps -a |grep -v grep |grep -c oscam >/dev/null
then
echo "OScam ok"
else
/usr/bin/camdctrl restart
fi

(my bin is called oscam)
-----------------------------------------------------------
This cron checks every 2 minutes?
cron

Quellcode

*/2 * * * * /usr/script/oscamcheck.sh

EverythingZen · 6. August 2012, 15:15

should work, yes.
and script is called every 2 minutes in the example.
if oscam emu is called oscam you have to substitude it in the sh file.

the best way to check if it's working is to stop oscam and look if it's started 2 minutes later.

eXaLtY · 6. August 2012, 19:35

lol

The script makes oscam restart every 2 minutes....
But the oscam was running!

EverythingZen · 6. August 2012, 20:58

Shell-Script

#!/bin/sh
if ps -a |grep -v grep |grep -c oscam >/dev/null
then
echo "OScam ok"
else
/usr/bin/oscam -b
fi

definetely works if you have done everything correct.

eXaLtY · 7. August 2012, 09:15

It's working! Thanks Dude.

Before I've made like this:

#!/bin/sh
while [ 1 ]
do
pidof oscam_1.00 >/dev/null
if [ $? -eq 0 ] ; then
echo "oscam_1.00 is running"
else
date=`date`
echo "Restarting oscam_1.00 : $date" >> /var/log/oscam_1.00.check
/var/bin/oscam &
if [ $? -eq 0 ] ; then
echo "oscam_1.00 is ok!"
else
echo "hm, didn't work. Try doing it manually"
fi
fi
sleep 20
done

now I've just use your code and it works!

How can I delete the log file before reaches an hudge amount of data?
I've tried to deleted manually but afterwords I doesn't create any more the log file.
The only solution was stop cron and started again, then a new log file was created in \cron\log

Maybe it's good to stick this Topic. It may help a lot of people!

Fred Bogus Trumper · 7. August 2012, 13:56

I think, there will be no problem with the amount of data. On my box there is a logfile in /tmp with more than 5000 lines and about 180kb - /tmp used is used 4% ...

an endless loop is not the best solution, because it can brake - or u run a watchdog for the watchdog

there are other solutions than the endless loop to check e.g every 20 seconds

One easy way for "autoclean" the logfiles:
If a defined number of lines (MAXLOGLINES=number) in the logfile is reached, move logfile to logfile.old. Then a new logfile will be created on next run. This solution creates only two files with max. 2 x $MAXLOGLINES lines

2nd way:
you overwirte /tmp/tmp/oscam_1.00.status on each run with the result of the last check and log only all restarts in /tmp/oscam_1.00.restarts. This will create less data. in the .status file u can see the last result and when the last check was done (wd still running?)

The numer of lines in /tmp/oscam_1.00.results could be cleaned on the same way as shown - if wanted

example (if all comment lines (# in front) are deleted, its very short - but do not delete the # in first line!)
/usr/script/oscam_watchdog.sh

Shell-Script

#!/bin/sh
# global variables
MAXLOGLINES=100
LOGFILE=/tmp/oscam_1.00.check
### get time
# function to get the correct system time in an indivudual format when called - output: yyyy-mm-dd@hh:mm:ss
now () {
date +%Y-%m-%d@%H:%M:%S
}
### logfile check
# function: if number of lines in logfile >= $MAXLOGLINES then move logfile
logfile_check () {
if [ $(wc -l $LOGFILE | awk '{print $1}') -ge $MAXLOGLINES ] ; then
mv -f $LOGFILE $LOGFILE.old
echo "$(now) - last $MAXLOGLINES lines of $LOGFILE moved to $LOGFILE.old" > $LOGFILE
fi
}
# if logfile exists, check number of lines with above function - if not: skip
[ -e $LOGFILE ] && logfile_check
### oscam check
# if result of "ps aux ...." = 0, oscam ist NOT running, else running
if [ $(ps aux | grep -v grep | grep -c "/usr/bin/oscam") = 0 ] ; then
echo "$(now) - oscam not running - restarting oscam_1.00" >> $LOGFILE
# alternative: write status in both files
# echo "$(now) - oscam not running - restarting oscam_1.00" >> /tmp/oscam_1.00.restarts
# echo "$(now) - oscam not running - restarting oscam_1.00" > /tmp/oscam_1.00.status
/usr/bin/oscam &
# may be better, because all oscam /tmp files created on start from NN2 start/stop script will cleaned/deleted:
# /usr/script/oscamscriptname.emu restart
# or 'camdctrl restart' if oscam only is used
else
echo "$(now) - oscam_1.00 is running ">> $LOGFILE
# alternative
# echo "$(now) - oscam_1.00 is running " > /tmp/oscam_1.00.status
fi
exit

Alles anzeigen

executed with cron every minute should be ok - but u can execute the script twice or tripple every minute with cron too:

crontab to execute every minute
*/1 * * * * /usr/script/oscam_watchdog.sh

executed twice within a minute (every 30 seconds)
*/1 * * * * /usr/script/oscam_watchdog.sh && sleep 30 && /usr/script/oscam_watchdog.sh

executed three times within a minute (every 20 seconds)
*/1 * * * * /usr/script/oscam_watchdog.sh && sleep 20 && /usr/script/oscam_watchdog.sh && sleep 20 && /usr/script/oscam_watchdog.sh

A third possibility beside cron and endless loop is a self-executing script which restarts itself after sleeping a defined time. But it is a little bit more comlicated to run correctly, in cause u have to prevent to start the script a second time when allready running

I think that the cron solution is the best and easiest way

have fun with testing & scripting

eXaLtY · 8. August 2012, 21:20

Once more you Rock dude!!!

I've noticed that the log file from cron gets bigger and bigger, this one needs to be flushed.....

I've used the other script mentioned and I saw that oscam is restarting even when was running.... this lead to the situation of not getting my cards to read at all and the proxies not able to reconnect due to an existing connection.

Any idea how to fix this?

Perhaps using your script will fix completely this situation, I will try tomorrow.

Another point is how to set a date by script when the dreambox is not connected to the sat cable?

I have this that it's currently working but maybe you guys know another way?

all_info=`wget -q -O /tmp/all.info islaremota.com/date.php?all=1; cat /tmp/all.info`
echo `date $all_info`

If the website doesn't respond I stay with no date and my box restarts at every 4 hours... lol

thks you all.

Cron - it's not working

Cron - it's not working

RE: Cron - it's not working

Shell-Script

Quellcode

Quellcode

Quellcode

Quellcode

Quellcode

Quellcode

Quellcode

Quellcode

Quellcode

Quellcode

Shell-Script

Quellcode

Shell-Script

Quellcode

Quellcode

Shell-Script

Shell-Script

Quellcode

Teilen