This morning I did what someone asked me to try to fix my echolink problem. remove temp files and logs.. So I removed all temp files in /tmp and backed up my logs in a directory and removed them from the /var/log/asterisk/ I then restarted the service with astres.sh It came back ok but I got this in verbose 3 [Jun 9 07:04:33] ERROR[32132]: chan_echolink.c:2432 do_el_directory: Error in directory download on naeast.echolink.org then looked in full log got nothing more then this.. [Jun 9 07:04:33] ERROR[32132] chan_echolink.c: Error in directory download on naeast.echolink.org So I went to the tmp file and downloaded the only file that was there. and it was the allstarlink node list.. so I went on and ping the naeast.echolink.org machine. cant reach it. from the allstar machine and from home.. Now If an error like that happen. do the echolink channel switch to another echolink server or it just die there? I did an echolink dbdump and the db was all screwed.. the first part of a line was not the node number, it was a truncked location.. to fix this I had to run the astres.sh script. So that mean that if in any of the partial db download an error happen, the db will get bad and it wont be fixed.. there must be a way to fix this.. one way that come to mind would be that if an echolink connection get rejected, the db is flush and a new full download happen.. that way if it was a corrupted db like I had the next time the user try to connect it will work, cause they always retry ;-) I will continu on reading that source code and will try to come with a fix..
On 06/09/2015 06:46 AM, pete M wrote:
This morning I did what someone asked me to try to fix my echolink problem.
remove temp files and logs..
So I removed all temp files in /tmp
and backed up my logs in a directory and removed them from the /var/log/asterisk/
I then restarted the service with astres.sh
It came back ok but I got this in verbose 3
[Jun 9 07:04:33] ERROR[32132]: chan_echolink.c:2432 do_el_directory: Error in directory download on naeast.echolink.org
then looked in full log got nothing more then this..
[Jun 9 07:04:33] ERROR[32132] chan_echolink.c: Error in directory download on naeast.echolink.org
There's more echolink servers than one. [el0] confmode=no call=W5OMR-R pwd=***** name=Geoff qth=Humble, TX email=ars.w5omr@gmail.com maxstns=08 rtcptimeout=10 node=4817 recfile=/tmp/recorded.gsm astnode=29656 context=radio-secure lat=0 lon=0 power=20 height=60 gain=13 dir=omni server1=server1.echolink.org server2=server3.echolink.org server3=server5.echolink.org
From: Geoff Sent: Tuesday, June 09, 2015 8:12 AM To: app_rpt-users@ohnosec.org Subject: Re: [App_rpt-users] echolink db update problem There's more echolink servers than one. _____________________________________________________ I do know that there are more then one. I was posting the state of the situation that caused the problem. It is way better to give the most info you got then to not tell something that could seem not usefull at first but prove to be the trigger of all this. here is my echolink setup _____________________________________________________ [el0] confmode=no call=VE2MRC-L pwd=************ name=Club ve2mrc qth=Montreal email=info@ve2mrc.com maxstns=20 rtcptimeout=10 node=464013 recfile=/tmp/recorded.gsm astnode=27821 context=radio-secure lat=45.557166 lon=-73.583084 power=9 height=3 gain=7 dir=8 server1=naeast.echolink.org server2=nasouth.echolink.org server3=servers.echolink.org #includeifexists custom/echolink.conf ______________________________________________________ So from what I can see it failed on server1 then went to server2 and downloaded a corrupted database.. Could it be that the buffer used to download on the first server was not flushed clean before downloading the new list from server 2? I know nothing on C. But I do know some basic, fortran pascal and cobol. and I think I could still code in Action (old atari 8 bit language). I did a whole BBS system with message database and game online using 2, atari 8600xl linked by there joystick port to exchange data from user connected to both computer by each there own modem.. I dont say i am a good programer.. But I did all this on my own with out any training.. I am not judging the work done here. I know that the echolink friver is not the prioroty here.. But this kind of bug could be fix by 2 method. First method, make a sanity check on the DB downloaded from echolink, this is a chalenge I know.. Or reach for a new DB when we see a reject of a user on echolink, cheap and dirty, but it would do the job.. Now are we looking to have a nice code or are we looking for a system that do work? I am not the one that will take that decision Jim your call!
participants (3)
-
Geoff -
pete M -
Pete M