Page 4 of 7
Posted: Thu Oct 12, 2006 1:03 pm
by zoneminder
Yep, just let me know the details and I will put something on it. Is it a build from source or a rpm?
How often does the box go doolally?
Posted: Thu Oct 12, 2006 1:06 pm
by jameswilson
ill pm em over to ya mate
well it was doing it all the time on the old distro, so i rebuilt it and took it back to site yesterday about 12:00 then at 17:00 ish it went funny. I restarted it and its been fine since. but i expect it to go again as it obvioulsy wasnt the old distro!
its rpm based mate
pm'ing now and thanks
The same to me
Posted: Fri Oct 13, 2006 4:33 am
by rcswrc
It's happening the same to me, only a system restart fix it, but the difference is a got a 206M and a 206 from axis, one at 7fps and the other at 5fps, and even worse it occurs every day at around 7:00 pm,
I'm running the latest 1.22.2 ZoneMinder version,
it says the same of buffer overrun, ant about the 1657121516.4(r), it happens some times.
I got a AMD64 Sempron 2600+ and 768MB RAM, I use a single 80GB sata hd only for the event folder.
I got something even stranger, cause it used to happen only with the 206M, so i deleted it and add it again as another and it started working ok, but then the same problem appeared on the 206, its too weird cause it's only with one camera.
i've also noticed that when that happens, the fps could get up to 100fps eventhough i set the max fps and max alarm fps to 7fps, and also the zma process of one camera gets about 65% CPU usage
For now my only solution is to restart the pc.
If you get something new please tell me!!!!!!!!!!!!!!!
Posted: Fri Oct 13, 2006 3:46 pm
by zoneminder
I have something running on James' machine now. Hopefully if/when it goes wrong it will tell us something at least.
Posted: Fri Oct 13, 2006 5:05 pm
by jameswilson
lol it aint missed a beat yet!! and caught a person on site who shouldnt have been while its playing up, not bad!!
rcswrc i thing the issues are diffrent and related to something that was fixed a while ago, you sure your on 1.22.2?
Posted: Sat Oct 14, 2006 6:28 pm
by rcswrc
yes i'm sure, I installed it a while ago, it's v1.22.2
you could watch it at zm.autopartesdelsur.com
with user test and password test
the one with the problem used to camara-error so i dont use it, but now its Atras.
it seems like zm only give me problems with one of the cameras, but it doesnt matter if its the 206 or the 206M
By the way, why do u think it's a different issue?
Posted: Sat Oct 14, 2006 9:45 pm
by jameswilson
mine affects all cameras and it looks like a timing issue somewhere. Plus you m,entioned the spike in fps, i dont have this. The reason your having a similar side effect is as your fps spikes it overlaods the proc, ending the event and starting a new one.
BTW what p[ath are you using for 206's. I had to use the mpeg path for the 206m bur prefer the jpeg path for the 206
Posted: Sat Oct 14, 2006 11:43 pm
by rcswrc
well u r right, i'll take your advice, and use the jpg path
thanks for your answer
Posted: Mon Oct 16, 2006 2:51 pm
by jameswilson
right then we have a development
All events have just gone when i have just checked in . Ill get the logs and let you know, i hope to f%%k nothing has happened i need fottage of as this is a system that has a bout 3 thefts a day on it
ill post the logs
Posted: Mon Oct 16, 2006 2:56 pm
by jameswilson
right then we dont have a development, i changed purge when full and didnt put it back right it removed everything as it should havem what a muppet!! i still hope nothing has happened
what's this?
Posted: Mon Oct 16, 2006 4:21 pm
by ManRoaming
What does this mean? I see these messageson my cams too.
Restarting capture daemon for Cam_20, shared memory not valid]
[/i]
Posted: Mon Oct 16, 2006 4:32 pm
by jameswilson
manroaming are you getting random ending of events?
if yes then post some more logs if not then please join a relevant one or start a new one
Posted: Wed Oct 18, 2006 9:36 am
by jameswilson
head up Phil, i have just had to restart it as its done it again
Posted: Wed Oct 18, 2006 10:22 am
by zoneminder
I've had a look. Unfortunately something had killed my monitoring script a few days ago so that doesn't have anything. We do still have the out of sequence times in the logs though which is odd.
I haven't been able to get to the bottom of the extremely short events yet but I have a suspicion about something else that precedes it and may be the cause.
For instance in your logs your have
Oct 18 10:11:09 id1461 zma_m7[3990]: INF [Cam_7: 286838 - Starting new event 370601]
Oct 18 10:11:14 id1461 zmaudit[4172]: INF [Filesystem event '7/370601' does not exist in database]
Oct 18 10:11:14 id1461 zma_m7[3990]: ERR [Can't open events/7/370601/013-capture.jpg: No such file or directory]
Oct 18 10:27:06 id1461 zmaudit[4172]: INF [Database event '7/370601' does not exist in filesystem]
I had thought that there may be an issue in this bit of code
Code: Select all
snprintf( sql, sizeof(sql), "insert into Events ( MonitorId, Name, StartTime, Width, Height, Cause, Notes ) values ( %d, 'New Event', '%s', %d, %d, '%s
', '%s' )", monitor->Id(), start_time_str, monitor->Width(), monitor->Height(), cause, text );
if ( mysql_query( &dbconn, sql ) )
{
Error(( "Can't insert event: %s", mysql_error( &dbconn ) ));
exit( mysql_errno( &dbconn ) );
}
id = mysql_insert_id( &dbconn );
and that as the events from several monitors are all created at once that the id reported may be that of an event created from a different monitor. This may still be the case in some circumstances so I will see if I can address that any way but in this case that is not the problem.
I am concentrating on zmaudit at the moment and have found something that might be a bit suspicious
Posted: Wed Oct 18, 2006 10:26 am
by jameswilson
Im assuming its hardware as i dont have it on any other systems and the other a8v board i have is dead. But as it doesnt kill it i assume it can be fixed. Do you need to do anything else with your monitoring script?