Old calls still in active list

I am making all sorts of statistics of inbound calls and I noticed some 15 minute blocks with huge call times. It turns out some calls get 'stuck' in the database for a really long time only to dissapear randomly (or it seems to be)
I think the actual Zap channel is not getting stuck, because I can see it appear again in the list, sometimes several times. I am running 2.0.4 with the agc directory replaced with the one Matt told me to to fix the 'old channel in your session' problem, but I have had the problem before that.
DB is in another machine, with 0.X load. This machines load is also 0.XX (0.32 at this time with 35 agents and a queue of 45 calls).
Is this a known problem? Should I just make a cron script to purge the DB of over 20 minutes calls? (average call is 3 minutes and over 15 minutes is rare)
Here is a sample:
VICIDIAL: Time On VDAD TRUNK SHORT: 0 2008-04-24 20:23:27
+---------------------+--------+--------------+--------------------+----------+---------+
| CHANNEL | STATUS | CAMPAIGN | PHONE NUMBER | CALLTIME | MINUTES |
+---------------------+--------+--------------+--------------------+----------+---------+
| Zap/4-1 | LIVE | TelmexTec | 56990515792 | 20:23:25 | 0:02 |
| Zap/21-1 | CLOSER | TelmexTec | 5645406424 | 20:23:19 | 0:08 |
| Zap/18-1 | CLOSER | TelmexTec | 5642222383 | 20:22:47 | 0:40 |
| Zap/7-1 | CLOSER | TelmexTec | 56968408830 | 20:22:41 | 0:46 |
| Zap/19-1 | CLOSER | TelmexCom | 56412984724 | 20:22:38 | 0:49 |
| Zap/45-1 | CLOSER | TelmexCom | 5625455242 | 20:21:56 | 1:31 |
| Zap/25-1 | CLOSER | TelmexTec | 5627455152 | 20:21:51 | 1:36 |
| Zap/14-1 | CLOSER | TelmexTec | 5625437558 | 20:21:41 | 1:46 |
| Zap/12-1 | CLOSER | TelmexTec | 5626873893 | 20:21:39 | 1:48 |
| Zap/30-1 | CLOSER | TelmexTec | 56977086928 | 20:21:37 | 1:50 |
| Zap/15-1 | CLOSER | TelmexTec | 5628521247 | 20:21:20 | 2:07 |
| Zap/23-1 | CLOSER | TelmexTec | 5625258865 | 20:20:57 | 2:30 |
| Zap/10-1 | CLOSER | TelmexTec | 56985880554 | 20:20:57 | 2:30 |
| Zap/26-1 | CLOSER | TelmexTec | 5628149592 | 20:20:50 | 2:37 |
| Zap/8-1 | CLOSER | TelmexTec | 5628129263 | 20:20:20 | 3:07 |
| Zap/2-1 | CLOSER | TelmexCom | 5622157796 | 20:20:18 | 3:09 |
| Zap/6-1 | CLOSER | TelmexTec | 5628213871 | 20:20:03 | 3:24 |
| Zap/40-1 | CLOSER | TelmexTec | 56Y04242016230 | 20:20:01 | 3:26 |
| Zap/17-1 | CLOSER | TelmexTec | 56Y04242017540 | 20:19:56 | 3:31 |
| Zap/24-1 | CLOSER | TelmexTec | 56997662994 | 20:19:51 | 3:36 |
| Zap/3-1 | CLOSER | TelmexTec | 5626494211 | 20:19:45 | 3:42 |
| Zap/9-1 | CLOSER | TelmexTec | 56977589937 | 20:19:38 | 3:49 |
| Zap/22-1 | CLOSER | TelmexCom | 56977064540 | 20:19:21 | 4:06 |
| Zap/13-1 | CLOSER | TelmexTec | 5622398192 | 20:19:08 | 4:19 |
| Zap/1-1 | CLOSER | TelmexTec | 56412987609 | 20:19:08 | 4:19 |
| Zap/31-1 | CLOSER | TelmexTec | 5625614995 | 20:19:08 | 4:19 |
| Zap/28-1 | CLOSER | TelmexTec | 5628440352 | 20:18:05 | 5:22 |
| Zap/20-1 | CLOSER | TelmexTec | 56998206636 | 20:15:56 | 7:31 |
| Zap/27-1 | CLOSER | TelmexTec | 5628600242 | 20:13:39 | 9:48 |
| Zap/29-1 | CLOSER | TelmexTec | 56322911311 | 20:11:47 | 11:40 |
| Zap/11-1 | CLOSER | TelmexCom | 5628599488 | 20:10:15 | 13:12 |
| Zap/5-1 | CLOSER | TelmexTec | 5625617830 | 20:04:56 | 18:31 |
| Zap/29-1 | CLOSER | TelmexCom | 5664231674 | 19:00:45 | 82:42 |
| Zap/23-1 | CLOSER | TelmexTec | 5643323124 | 17:06:29 | 196:58 |
| Zap/19-1 | CLOSER | TelmexTec | 5624493971 | 15:46:54 | 276:33 |
| Zap/6-1 | CLOSER | TelmexTec | 5625852663 | 15:26:55 | 296:32 |
| Zap/19-1 | CLOSER | TelmexTec | 5633716391 | 13:10:18 | 433:09 |
+---------------------+--------+--------------+--------------------+----------+---------+
37 calls being placed on server 192.168.0.13
I think the actual Zap channel is not getting stuck, because I can see it appear again in the list, sometimes several times. I am running 2.0.4 with the agc directory replaced with the one Matt told me to to fix the 'old channel in your session' problem, but I have had the problem before that.
DB is in another machine, with 0.X load. This machines load is also 0.XX (0.32 at this time with 35 agents and a queue of 45 calls).
Is this a known problem? Should I just make a cron script to purge the DB of over 20 minutes calls? (average call is 3 minutes and over 15 minutes is rare)
Here is a sample:
VICIDIAL: Time On VDAD TRUNK SHORT: 0 2008-04-24 20:23:27
+---------------------+--------+--------------+--------------------+----------+---------+
| CHANNEL | STATUS | CAMPAIGN | PHONE NUMBER | CALLTIME | MINUTES |
+---------------------+--------+--------------+--------------------+----------+---------+
| Zap/4-1 | LIVE | TelmexTec | 56990515792 | 20:23:25 | 0:02 |
| Zap/21-1 | CLOSER | TelmexTec | 5645406424 | 20:23:19 | 0:08 |
| Zap/18-1 | CLOSER | TelmexTec | 5642222383 | 20:22:47 | 0:40 |
| Zap/7-1 | CLOSER | TelmexTec | 56968408830 | 20:22:41 | 0:46 |
| Zap/19-1 | CLOSER | TelmexCom | 56412984724 | 20:22:38 | 0:49 |
| Zap/45-1 | CLOSER | TelmexCom | 5625455242 | 20:21:56 | 1:31 |
| Zap/25-1 | CLOSER | TelmexTec | 5627455152 | 20:21:51 | 1:36 |
| Zap/14-1 | CLOSER | TelmexTec | 5625437558 | 20:21:41 | 1:46 |
| Zap/12-1 | CLOSER | TelmexTec | 5626873893 | 20:21:39 | 1:48 |
| Zap/30-1 | CLOSER | TelmexTec | 56977086928 | 20:21:37 | 1:50 |
| Zap/15-1 | CLOSER | TelmexTec | 5628521247 | 20:21:20 | 2:07 |
| Zap/23-1 | CLOSER | TelmexTec | 5625258865 | 20:20:57 | 2:30 |
| Zap/10-1 | CLOSER | TelmexTec | 56985880554 | 20:20:57 | 2:30 |
| Zap/26-1 | CLOSER | TelmexTec | 5628149592 | 20:20:50 | 2:37 |
| Zap/8-1 | CLOSER | TelmexTec | 5628129263 | 20:20:20 | 3:07 |
| Zap/2-1 | CLOSER | TelmexCom | 5622157796 | 20:20:18 | 3:09 |
| Zap/6-1 | CLOSER | TelmexTec | 5628213871 | 20:20:03 | 3:24 |
| Zap/40-1 | CLOSER | TelmexTec | 56Y04242016230 | 20:20:01 | 3:26 |
| Zap/17-1 | CLOSER | TelmexTec | 56Y04242017540 | 20:19:56 | 3:31 |
| Zap/24-1 | CLOSER | TelmexTec | 56997662994 | 20:19:51 | 3:36 |
| Zap/3-1 | CLOSER | TelmexTec | 5626494211 | 20:19:45 | 3:42 |
| Zap/9-1 | CLOSER | TelmexTec | 56977589937 | 20:19:38 | 3:49 |
| Zap/22-1 | CLOSER | TelmexCom | 56977064540 | 20:19:21 | 4:06 |
| Zap/13-1 | CLOSER | TelmexTec | 5622398192 | 20:19:08 | 4:19 |
| Zap/1-1 | CLOSER | TelmexTec | 56412987609 | 20:19:08 | 4:19 |
| Zap/31-1 | CLOSER | TelmexTec | 5625614995 | 20:19:08 | 4:19 |
| Zap/28-1 | CLOSER | TelmexTec | 5628440352 | 20:18:05 | 5:22 |
| Zap/20-1 | CLOSER | TelmexTec | 56998206636 | 20:15:56 | 7:31 |
| Zap/27-1 | CLOSER | TelmexTec | 5628600242 | 20:13:39 | 9:48 |
| Zap/29-1 | CLOSER | TelmexTec | 56322911311 | 20:11:47 | 11:40 |
| Zap/11-1 | CLOSER | TelmexCom | 5628599488 | 20:10:15 | 13:12 |
| Zap/5-1 | CLOSER | TelmexTec | 5625617830 | 20:04:56 | 18:31 |
| Zap/29-1 | CLOSER | TelmexCom | 5664231674 | 19:00:45 | 82:42 |
| Zap/23-1 | CLOSER | TelmexTec | 5643323124 | 17:06:29 | 196:58 |
| Zap/19-1 | CLOSER | TelmexTec | 5624493971 | 15:46:54 | 276:33 |
| Zap/6-1 | CLOSER | TelmexTec | 5625852663 | 15:26:55 | 296:32 |
| Zap/19-1 | CLOSER | TelmexTec | 5633716391 | 13:10:18 | 433:09 |
+---------------------+--------+--------------+--------------------+----------+---------+
37 calls being placed on server 192.168.0.13