Almost completed - ZM, Spectra8, Fedora Core 3

[zoneminder]

I'm on FC1 with kernel 2.4.22 and stock bttv drivers I think.

My modprobe.conf contains the following lines

alias char-major-81-2 bttv
alias char-major-81-3 bttv
alias char-major-81-4 bttv
alias char-major-81-5 bttv
options bttv card=94,94,94,94 pll=1
alias char-major-81-6 bttv
alias char-major-81-7 bttv
alias char-major-81-8 bttv
alias char-major-81-9 bttv
options bttv card=42,42,42,42 pll=1 radio=0 tuner=4 sloppy=1 chroma_agc=1 mmap=1 vbibufs=4 v4l2=1 gbuffers=16

This is for two cards, the Spectra and a LML one. I forget which config applies to which but if yours doesn't match either then you could try them both and see which one works. I've never had a reboot at all on this box, except when I've asked for one and I've done some pretty nasty to things to the cards while developing ZM as you can imagine

Phil

[somepdguy]

Hi all,

OK sorry - i have whipped out the card and hard disk out and put them in another machine and it works without rebooting so it must be an issue with the shuttle unit I was using.

Sorry, a few years ago it would have been the first thing I would have tried but I am getting to old for all this

Thanks again for the help, the spectra 8 card seems OK for me now using

modprobe bttv card=42,42,42,42 pll=1 radio=0 tuner=4 sloppy=1 chroma_agc=1 mmap=1 vbibufs=4 v4l2=1 gbuffers=16

if I run into any more problems I will keep the thread updated.

Thanks for a great program again Phil.


Pete

[pmurdock]

Okay just to provide further updates on my setup.

I had everything working fine, and then noticed having dropped events.. About 2 every hour.. So I would have in the database the

9:00
9:10 - DROPPED or DELETED
9:20
9:30
9:40
9:50 - DROPPED or DELETED
10:00

I've been through the previous threads about this issue, and in the end I just wiped out the mysql database and started over again with that. It seems to have fixed the issue. I'm back in "mocord" mode and it doesn't seem to be skipping events. Although I have NOT implemented PurgeWhenFull yet.. Somehow I think it has to do with this filter.

On another note, all the old video footage that I have I would like to delete and doing it manually in linux is SLOW as molasses ... at least the way I do it which is

cd /var/lib/events/1
yes | rm -R *

then I tried

rm -f -R *


this takes FOREVER to clean out the directories with the images.. Is there another more efficient way to wipe out all those jpg's?

cheers,
Paul

[cordel]

On the issue with purgeWhenFull,
Phil is there a way to query what the most recent event is or could be? If so we could use something like this equation to may remidy it.

most resent event - (2 * n)

n being number of monitors defined and in a record mode.

So that would give something to base as a protected section of events that may be in transition plus one just in case.

[zoneminder]

The easiest way to delete events is to remove the record from the DB and have zmaudit.pl clean up after you. So in mysql do

Code: Select all

delete from Events where Id < 1000;

or however you want to phrase it.

Corey, I'm not sure what you are try to suggest. That the most recent events should automatically be left alone so they don't get deleted by accident? Have I got that right? In theory no event less than 5 minutes old should be touched anyway, but for some reason some get wiped. I wish I could figure out why!

Cheers

Phil

[pmurdock]

Even without enabling PurgeWhenFull filter I lose events!

This is "record" mode..

Contents of zmaudit.log (when problems start happening)

Code: Select all

Deleting 1 old images
Found database monitor '1', got 23 events
Found filesystem monitor '1', got 23 events
Database event '1/23' does not exist in filesystem, deleting
Found database monitor '1', got 24 events
Found filesystem monitor '1', got 25 events
Filesystem event '1/23' does not exist in database, deleting
Found orphaned frame records for event '23', deleting
Found database monitor '1', got 25 events
Found filesystem monitor '1', got 25 events
Database event '1/26' does not exist in filesystem, deleting
Found database monitor '1', got 26 events
Found filesystem monitor '1', got 27 events
Filesystem event '1/26' does not exist in database, deleting
Found orphaned frame records for event '26', deleting
Found database monitor '1', got 27 events
Found filesystem monitor '1', got 27 events
Found database monitor '1', got 29 events
Found filesystem monitor '1', got 29 events
Found database monitor '1', got 30 events
Found filesystem monitor '1', got 30 events
Database event '1/32' does not exist in filesystem, deleting
Found database monitor '1', got 31 events
Found filesystem monitor '1', got 32 events
Filesystem event '1/32' does not exist in database, deleting
Found orphaned frame records for event '32', deleting
Found database monitor '1', got 32 events
Found filesystem monitor '1', got 32 events
Database event '1/35' does not exist in filesystem, deleting
Found database monitor '1', got 33 events
Found filesystem monitor '1', got 34 events
Filesystem event '1/35' does not exist in database, deleting
Found orphaned frame records for event '35', deleting
Found database monitor '1', got 34 events
Found filesystem monitor '1', got 34 events
Database event '1/38' does not exist in filesystem, deleting
Found database monitor '1', got 35 events
Found filesystem monitor '1', got 36 events
Filesystem event '1/38' does not exist in database, deleting
Found orphaned frame records for event '38', deleting
Found database monitor '1', got 36 events
Found filesystem monitor '1', got 36 events
Database event '1/41' does not exist in filesystem, deleting
Found database monitor '1', got 37 events
Found filesystem monitor '1', got 38 events
Filesystem event '1/41' does not exist in database, deleting
Found orphaned frame records for event '41', deleting

Contents of zmwatch.log

Code: Select all

Can't execute:  at /usr/lib/zm/bin/zmwatch.pl line 121.
Watchdog starting at 05/06/12 19:25:49
Watchdog pausing for 30 seconds
Watchdog starting at 05/06/12 19:35:43
Watchdog pausing for 30 seconds
Can't execute:  at /usr/lib/zm/bin/zmwatch.pl line 121.
Watchdog starting at 05/06/12 20:12:04
Watchdog pausing for 30 seconds
Watchdog starting at 05/06/12 20:18:37
Watchdog pausing for 30 seconds
Watchdog starting at 05/06/13 08:55:00
Watchdog pausing for 30 seconds
Can't execute:  at /usr/lib/zm/bin/zmwatch.pl line 121.
Watchdog starting at 05/06/14 22:27:29
Watchdog pausing for 30 seconds
Watchdog starting at 05/06/15 12:08:25
Watchdog pausing for 30 seconds

another curiousity.. issues with zmwatch.pl for some reason... a known issue?

[zoneminder]

Now this is starting to p*ss me off!! I need to get to the bottom of why one minute zmaudit can find the db event and not the filesystem one and the next it's the other way around! If you get this consistently it might be helpful if I can send you a debug version of that script so I can figure out what's going on.

The error from zmwatch is odd as well. Basically it's just rerunning a very simple query and seems to be getting an error result with no error code. Very odd indeed. Have you checked your DB for errors?

Phil

[zoneminder]

I've been tinkering around on your system and adding some debug to your zmaudit.pl. After a while I realised that it wasn't the latest version so I've put the 1.21.1 zmaudit.pl on there and 'so far' not seen any unusually deleted event. Maybe you can keep an eye on the log file for the next few days to confirm.

Phil

[cordel]

So does this mean that the wrong version of zmaudit was in the tar ball or that it didn't update correctly

[pmurdock]

looks good so far.. I haven't noticed any dropped events..

so perhaps an outdated zmaudit.pl made it into the packaged fedora core 3 RPM's..


cheers!

Paul

[caseystone]

in my (brief) experience, unplugging a camera that is set to anything but Function None WILL crash ZMC (or one of the key processes). That also occasionally happens to me just with normall running. Not sure why.. I'll log in and see that the load is at like 14 or something, when it's normally under 1.

[cordel]

Hi caseystone,
You might want to start a new thread with this since it does not releate to any issues here and, elaborate a bit more.
Regards,
Cordel

[caseystone]

Ha! That's really weird. I wrote that in response to another thread I was reading.. It did relate to that one, though it was kind of a stale thread. I'm not sure how it ended up in this thread.

-Casey