Strange timeout behaviour

Hi all,

I'm seeing unusual timeout behaviour. I have private ASL3 node 1717 connected to 618035 hub, which is connected to 59606 where our weekly news broadcast originates.
I have a cron entry to disable timeout on 1717 so that the news plays for the hour duration.

# Sunday morning broadcast
45 08 * * 7 /usr/sbin/asterisk -x "rpt cmd 1717 cop 8"
15 09 * * 7 /usr/sbin/asterisk -x "rpt cmd 1717 cop 7"

Sometimes the cron job fails to disable timeout. This morning timeout was successfully disabled at 8:45 ready for the 9:00 start. The broadcast started, but after about 20 minutes 1717 timed out. "rpt stats 1717" revealed that timeout was reactivated.

Why is my node behaving like this? I have seen similar behaviour with extended hang time that I enable for a weekly net. Sometimes it activates, sometimes it doesnt, and sometimes when it does activate it resets itself in the same way the timeout does.

Thanks,
Dion.

Would you consider testing it's activation in the asl system scheduler ?
See if you get the same results.

Perhaps you have a string somewhere that is tuning it back on that is set for the wrong time. Or set to wrong macro string to be executed.
Plenty of possibilities that you are going to have to search out.

Obviously, it is getting the command to turn back on, you just need to figure out where that is coming from.

To be clear, could this be a 'link timeout' you are hearing which is a separate setting/timer.
Not the same as a TX timeout. Which is what cop 7/8 deal with.

Hi Mike,

I have no macros enabled in rpt.conf
There is nothing else in cron that should cause the timeout to re-enable during the broadcast period, see below.
There was no DTMF received during the broadcast, 1717 is a simplex node.

# m h  dom mon dow   command


# Sunday morning broadcast
45 08 * * 0 /usr/sbin/asterisk -x "rpt cmd 1717 cop 8"
15 09 * * 0 /usr/sbin/asterisk -x "rpt cmd 1717 cop 7"

# Tuesday night rebroadcast
45 19 * * 2 /usr/sbin/asterisk -x "rpt cmd 1717 cop 8"
15 21 * * 2 /usr/sbin/asterisk -x "rpt cmd 1717 cop 7"

# RAOTC
45 19 */100,1-7 * MON /usr/sbin/asterisk -x "rpt cmd 1717 cop 8"
45 20 */100,1-7 * MON /usr/sbin/asterisk -x "rpt cmd 1717 cop 7"

Its a fresh ASL3 image on a Pi 3B+ with only basic config changes to make it work.

Thanks,
Dion.

It looks like you pasted in the content of your crontab file. Was this using :

crontab -e (or crontab -l)

OR were you using :

sudo crontab -e (or sudo crontab -l)

I'm not sure about using 08 and 09 as hourly values. One would hope the leading zero's don't present an issue ... but maybe try just 8 and 9.

Lastly, what's a "day of month (0-31)" value of */100,1-7 supposed to mean?

Dion, are you sure this is a TX timeout and not a link activity timeout ?

If you don't understand the difference, then please say so as I will not continue until I know that part for sure. It would be a waste of both our time.

I'm logged in as root, I do everything as root, its roots cron.

crontab.guru has no problems with leading zeros on the hour values. I did put the zeros in there just as a test to see if that was more reliable.

Apparently */100,1-7 * MON is a hack to trigger an action on the first Monday of the month only, which is when the RAOTC broadcast plays. “At 19:45 on every 100th day-of-month and every day-of-month from 1 through 7 if it's on Monday.”

This is a TX timeout, not a link activity timeout. I have no link activity timers configured.
The TX timeout being enabled is evident by the output of rpt stats 1717 producing

Time out timer...................................: ENABLED

Rather than

Time out timer...................................: DISABLED

This being controlled by cop 7/8, with no control inputs, no macros, no scripts, no cron jobs, I dont understand how it can revert back to being enabled seemingly randomly by itself.

If you manually set the TOT off with dtmf, does it stay off ?

0 0 1-7 * 1

That is the first monday of the month in cron.

Or this also works...
0 0 * * MON#1

So the last two Sundays the same thing happened. Before the broadcast started the timeout was disabled. The broadcast transmit for approximately 17 minutes then timed out. Checking the stats showed that the timeout was enabled again. This has got me beat..

I guess I have to ask again since you did not answer the first time.

cop,45 ; Link Activity timer enable -LATENA
cop,46 ; Link Activity timer disable -LATDIS

lnkacttime=1800

;945 = cop,45 ; Link Activity timer enable
;945 = cop,46 ; Link Activity timer disable

;lnkactenable = 0 ; Set to 1 to enable the link activity timer. Applicable to standard nodes only.

;lnkacttime = 1800 ; Link activity timer time in seconds.

All the relevant link activity timer settings are commented out.

I think you should run the ‘cop 46’ from the command line once and be sure it’s not been enabled by mistake. Commented out they only can not be run by dtmf.