Hi,

I do see this in logs and this seems normal, cause we do store that $sht and in a case if call is cancelled or ended we do not clean that $sht until it resets in 120 seconds, so if any register arrives without call during that 120 seconds, this is what you will see. Let assume somebody calls a subscriber, then we do store transaction and  we do this "$sht(vtp=>stored::$rU) = 1", if call ends and by some reason without any other call another register arrives from subscriber with same "$tU" within 120 seconds after ($sht(vtp=>stored::$rU) = 1;), then this register will go to PUSHJOIN and we check if transaction is stored, because there is no call it is not stored and then it checks if $sht(vtp=>stored::$rU) equal to 1, because this is not resetted yet(120 seconds do not pass), ts_append tries to append branch, but there is no branch and that why you see this error.

I would like to add that for me this error seems harmless and everything works flawlessly.

Jurijs


On Wed, Nov 25, 2020 at 2:39 PM Anthony Alba <ascanio.alba7@gmail.com> wrote:
Do you ever see the following in your logs in route[REGISTER] from the
callee user?
I think this happens when
1. All transactions have been cleaned up by tsilo, so nothing to append
2. $sht(vtp=>stored::$tU) has not timed out in the 120s
3. and $tU/callee also just happens to  re-REGISTER before
$sht(vtp=>stored::$tU) is removed

I guess it is harmless but it is an eyesore to my logs manager, any
clever tricks

I think this code logic
if ($var(hjoin)==0)
{
if ($var(hstored))
ts_append("location", "$tU");
return;
}
causes the REGISTER of the callee to happen to try to ts_append().

2020-11-25T20:30:01+08:00 127.0.0.1 tsilo.ts_append()
sip:charles@voice.example.com
2020-11-25T20:30:01+08:00 127.0.0.1 ERROR: tsilo [ts_append.c:64]:
ts_append(): failed to retrieve record for
sip:charles@voice.zenquark.com

_______________________________________________
Kamailio (SER) - Users Mailing List
sr-users@lists.kamailio.org
https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users