Debian packages

[alexo]

Hello,
Can anyone provide 1.30.2 Debian packages.
I want to try 1.30.2 on Debian 8 (Jessie).
Currently running 1.29.0 but experiencing a number of issues (the worst one is zoneminder service random crashes - 1-2 times a day)
Thanks

[bbunge]

The latest ZM for Debian that I know of is 1.30.0 on Debian 9. If you are having stability issues upgrading likely would not help. Lower your camera resolution and frame rate. Make sure you have space in tmpfs.

[alexo]

I have 8 cams running at 2fps.
SHM set to 4G, the system has 8GB RAM
/tmp was increased and never goes above 50%
tmpfs 4.0G 2.7G 1.4G 68% /dev/shm
tmp 984M 403M 534M 43% /tmp

Here is part of the logs(during crash):

May 3 18:16:36 systemd[1]: zoneminder.service: main process exited, code=exited, status=255/n/a
May 3 18:16:40 systemd[1]: zoneminder.service: control process exited, code=exited status=1
May 3 18:16:51 systemd[1]: Unit zoneminder.service entered failed state.


During that time, I can find the following in the components logs:

/var/log/zm #grep "18:16" *
zmdc.log:05/03/2017 18:16:20.092607 zmdc[20196].INF ['zma -m 12' sending stop to pid 20462 at 17/05/03 18:16:20]
zmdc.log:05/03/2017 18:16:26.093303 zmdc[20196].WAR ['zma -m 12' has not stopped at 17/05/03 18:16:26. Sending KILL to pid 20462]
zmdc.log:05/03/2017 18:16:36.724093 zmdc[20196].INF ['zma -m 12' exited normally]
zmfilter.log:04/30/2017 15:18:16.986772 zmfilter[9289].INF [Deleting event 20529 from Monitor 9]
zmpkg.log:05/03/2017 18:16:37.287845 zmpkg[31658].INF [Sanity checking States table...]
zmpkg.log:05/03/2017 18:16:37.288264 zmpkg[31658].INF [Command: stop]
zmwatch.log:05/03/2017 18:16:19.950236 zmwatch[20419].INF [Analysis daemon for PoleBackyard needs restarting, time since last analysis 6 seconds (1493824579-1493824573)]
zmwatch.log:05/03/2017 18:16:19.950329 zmwatch[20419].INF [Restarting analysis daemon for PoleBackyard]

Any other ideas to solve this issue?
Thanks

[bbunge]

Looks like you are not running out of resources and 2 fps is OK. I suspect you are using a low resolution? If you are building up a lot of events it may be possible your MySQL is running out of resources. Install and run mysqltuner. You may have to increase innodb_buffer_pool_size. Might also check the number of connections. I am running 17 cameras on an i5 Quad with 8 GIG ram, 2 GIG /dev/shm, 2 to 4 fps, 320x240 32 bit color. Am keeping about 60K events, have the innodb_buffer_pool_size set to 1G, 1000 connections as there are at least 6 PC's watching montage at any given time. I recently upped the RAM to 16 GIG as I was trying to keep Ubuntu 16.04 from using swap.

[SlowScan]

http://www.deb-multimedia.org has 1.30.2.

If you are not familiar with deb-multimedia. org, which is not officially sanctioned by Debian, please read https://wiki.debian.org/DebianMultimedia/FAQ before using.

[alexo]

Well the CPUs run at 100% since 6 of the cams are recording at 1080px and the cpus are a bit old (2 Xeon cores 3060 @ 2.40GHz). Constant recording and no one watching montage.I use zmu to dump all images each 5 seconds to an apache web page.
I have only ~2000 events (2 days of full recording) and I don't think MySQL is the bottleneck.

Expecting 2 new servers so I'll soon virtualize this zoneminder instance (it will run on 2x6cores dell hardware) but I was hoping to fix this one meanwhile (or at least figure it out)

I'll check as well deb-multimedia.org ...

Thanks

[bbunge]

Well the CPUs run at 100% since 6 of the cams are recording at 1080px and the cpus are a bit old (2 Xeon cores 3060 @ 2.40GHz). Constant recording and no one watching montage.I use zmu to dump all images each 5 seconds to an apache web page.
I have only ~2000 events (2 days of full recording) and I don't think MySQL is the bottleneck.

Expecting 2 new servers so I'll soon virtualize this zoneminder instance (it will run on 2x6cores dell hardware) but I was hoping to fix this one meanwhile (or at least figure it out)

I'll check as well deb-multimedia.org ...

Thanks

I feel you would do better not running ZM on a virtual machine or in a container. You could also lower the resolution on those 6 cameras by half. This is a security system not a high res movie.

[knight-of-ni]

CPU's at 100% is guaranteed to be the reason zoneminder stops on you. You can't expect to run any system like that.

Even for old hardware, your load still sounds a little high for 2 fps. Double check you are setting the frame rate in the camera, not in zoneminder.
Leave the Max fps fields in zoneminder empty.

Cherry picking events out of your logs is not particularly helpful for a problem like this. We need to see a lot of context to understand better what is happening. If you still have issues, show at least one unedited page of events from the zoneminder log. The more pages the better. Link to a Gist or pastebin if you need to.

[alexo]

CPUs are averaging 65-70% but system load is constantly above 2.
FPS is set on all cameras and nothing changed in ZM.
As I do recording only, I guess could save a lot of resources If I could directly store in h246