Fun with old hardware

nombrandue · Post by **nombrandue** » Wed Nov 07, 2007 4:13 am

Ok, so it isn't ALL old hardware. I have this system doing a bit, but it was supposed to be a DVR security system. I am having problems with Zoneminder in the fact that anything over the top kills it. What I mean by this, is if I go above 100x100 pix. Gray color, or change the format type (to PAL or NTSC) to anything but Auto, I get tons of problems in the logs, errors, crash loops.

Error log (zmdc.log)
11/06/2007 20:56:38.673546 zmdc[3125].INF ['zmc -d /dev/video0' starting at 07/11/06 20:56:38, pid = 15988]
11/06/2007 20:56:39.123747 zmdc[3125].ERR ['zmc -d /dev/video0' exited abnormally, exit status 255]
11/06/2007 20:56:42.125962 zmdc[3125].INF ['zmc -d /dev/video0' starting at 07/11/06 20:56:42, pid = 16008]
11/06/2007 20:56:42.524787 zmdc[3125].INF ['zmc -d /dev/video0' stopping at 07/11/06 20:56:42]
11/06/2007 20:56:42.537354 zmdc[3125].INF ['zmc -d /dev/video0' crashed, signal 14]
11/06/2007 20:56:46.115655 zmdc[3125].INF ['zmc -d /dev/video0' starting at 07/11/06 20:56:46, pid = 16017]
11/06/2007 20:56:46.517928 zmdc[3125].ERR ['zmc -d /dev/video0' exited abnormally, exit status 255]
11/06/2007 20:56:46.521079 zmdc[3125].INF [Starting pending process, zmc -d /dev/video0]
11/06/2007 20:56:46.527583 zmdc[3125].INF ['zmc -d /dev/video0' starting at 07/11/06 20:56:46, pid = 16020]
11/06/2007 20:56:46.818008 zmdc[3125].ERR ['zmc -d /dev/video0' exited abnormally, exit status 255]
11/06/2007 20:56:47.931790 zmdc[3125].WAR [Can't find process with command of 'zmtrack.pl -m 19']
11/06/2007 20:56:49.303880 zmdc[3125].INF [Command 'zma -m 19' removed from pending list at 07/11/06 20:56:49]
11/06/2007 20:56:51.004496 zmdc[3125].INF [Starting pending process, zmc -d /dev/video0]
11/06/2007 20:56:51.012350 zmdc[3125].INF ['zmc -d /dev/video0' starting at 07/11/06 20:56:51, pid = 16025]

11/06/2007 21:55:49.031529 zmdc[3125].INF [Starting pending process, zmc -d /dev/video]
11/06/2007 21:55:49.038277 zmdc[3125].INF ['zmc -d /dev/video' starting at 07/11/06 21:55:49, pid = 18406]
11/06/2007 21:55:49.149755 zmdc[3125].ERR ['zmc -d /dev/video' exited abnormally, exit status 255]

Of course these logs are truncated and pruned a bit. Now, my hardware seems to be the error (the zm proccesses suck up a lot of cycles)

the hardware is:
1.00 Ghz AMD K7 CPU
Abits KT7RAID board
512 Meg RAM
Apollo Digital Surveillance Model W-DVR
a rather archaic All in Wonder card
Voodoo Banshee 16 video card.

Save for the Apollo card everything is odds and ends. I am wondering, would it be better just to scrap the old system as my DVR, and get newer equipment for it, or is there something I can do to get this to play well with the cobweb collectors I have put together for this?

Lee Sharp · Post by **Lee Sharp** » Wed Nov 07, 2007 1:42 pm

I think it is the low ram. I have never tried anything less than a gig, and that can fill up. The low res, and black and white only support this.

nombrandue · Post by **nombrandue** » Wed Nov 07, 2007 2:20 pm

Ok, that makes sense. I just need to find some better sticks to plug into this system and try that. That would explain everything going red when I change to a setting it doesn't like.

I will get some more memory into that and see if that changes it. Also is there a decent list of what the codes in the errors are? I think that would make it a little easier to find out what may be the problem is, or at least point in the right direction. So far I have seen error numbers 6 and 255.

agoodm · Post by **agoodm** » Thu Nov 08, 2007 3:48 pm

I run 4 cameras at cif resolution 12 fps 100 frames ring buffer on each camera on 512mb ram.

nombrandue · Post by **nombrandue** » Sat Nov 17, 2007 10:50 pm

Ok, I have, after a bit of playing, got an RBG24 100x100 image that runs at 4FPS on the camera I have running. anything bigger than 100x100 will go red, apparently the memory is not able to handle that. Now, what I am wondering, aside form camera constraints, what is the max FPS I can get from 1.5 GB/s of RAM?

nombrandue · Post by **nombrandue** » Sun Nov 18, 2007 4:36 am

Ok, I have adjusted it, definately a memory limit (lower ring lead to a higher image size) and have been playing with different settings. The Camera I am testing with is in my hallway/foyer of my apartment, and the carpet outside of the front door where the camera is looking at, is brown pattered. The Cam shows it as blue. As well I don't have exact specs of the camera, at least none that directly match anything I have read so far on ZM's site. It is a luxor LX-512SH.

I just am wondering, if there is a way I can tune/play with it in ZM to tune the color right, or if it is a cam issue and I just say forget it? Though it is amusing, because I prefer the blue to the brown...

nombrandue · Post by **nombrandue** » Mon Nov 19, 2007 2:07 am

Ok, another decent update after a lot of playing around with the settings on the low end system. Running the Camera at 240x180 RBG24 seems to eat a lot of memory and not free it up outside of the buffer. It runs from 10-50 minutes and then slowly dies. It will start to flicker (brief delayed reloads that got from the image to blank back to the image) and finally bales out to not showing any images. If I restart (stop and start instead of restart) zone Minder, and it works fine for another period. I will be getting more memory later this week to see how this works in more than 512 ram. As one of the posts earlier mentioned, I lowered my ring buffer to the lowest settings I could, and it worked a little more reliably for a while.

I guess part of the issue is that I have a few other core systems running on the system.

nombrandue · Post by **nombrandue** » Mon Nov 19, 2007 9:35 pm

One good question to tag along with my string of issues:

Is anyone using squid as a reverse proxy in front of zone minder? If so, have you had issues with streaming images showing up when viewing ZM though the proxy? I am currently unable to view any of the data from outside of my local network, due to this proxy and images not being shown. Aside from removing it from the proxy, is there a good solution?

nombrandue · Post by **nombrandue** » Wed Nov 21, 2007 11:23 pm

Ok, Memory in the box has been maxed out. I am still seeing the crashes of zmdc on my video device. Also I am seeing a lot of this crap:

Nov 21 17:20:01 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:01 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:01 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:01 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:02 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:02 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:02 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:02 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:03 s_dgram@minazuki.tsukinokage.net zmu[5306]: INF [New Debug Level = 9, New Debug Log = /var/log/zoneminder/debugging.log]
Nov 21 17:20:03 s_dgram@minazuki.tsukinokage.net zmu[5306]: INF [Authenticated user 'nombrandue']
Nov 21 17:20:03 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:03 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:03 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:03 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:03 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:04 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:04 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:04 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:04 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:05 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:05 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:05 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:05 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:05 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:06 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:06 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:06 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:06 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:06 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:07 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:07 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:07 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:07 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:08 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:08 s_dgram@minazuki.tsukinokage.net zmu[5310]: INF [New Debug Level = 9, New Debug Log = /var/log/zoneminder/debugging.log]
Nov 21 17:20:08 s_dgram@minazuki.tsukinokage.net zmu[5310]: INF [Authenticated user 'nombrandue']
Nov 21 17:20:08 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:08 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:08 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:09 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:09 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:09 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:09 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:09 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:10 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:10 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:10 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:10 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:11 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:11 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:11 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:11 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:11 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:12 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:12 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:12 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:12 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:12 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:13 s_dgram@minazuki.tsukinokage.net zmu[5313]: INF [New Debug Level = 9, New Debug Log = /var/log/zoneminder/debugging.log]
Nov 21 17:20:13 s_dgram@minazuki.tsukinokage.net zmu[5313]: INF [Authenticated user 'nombrandue']
Nov 21 17:20:13 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:13 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:13 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]
Nov 21 17:20:13 s_dgram@minazuki.tsukinokage.net zms[5188]: INF [Caution: quantization tables are too coarse for baseline JPEG]

Now, I have YET to find anything that explains what any of these errors mean (Just what the hell does the error status of 255, or 9 mean in relation to zmdc and zmtrigger respectively?)

All in all I am rather dissatisfied with this program right now, just because, without reading the entire source line by line, I haven't a clue as to what the hell it is trying, and dying to do.

jameswilson · Post by **jameswilson** » Thu Nov 22, 2007 12:08 am

Sorry you feel that way but it works fine on my and other peoples hardware.

I will say that i run 2 cams myself on an old 750 meg duron with 512meg and no problems for about 3 yrs. But im using the live cd and most people that have problems dont.

Just guessing mind

nombrandue · Post by **nombrandue** » Thu Nov 22, 2007 12:55 am

Well, I even removed most of my install and installed the Fedora 7 RPM from YUM. I removed one of the capture cards (a rather unused ATI allinwonder card) and left the Apollo in. I get this from zmu:

Video Capabilities
Name: BT878 video ( *** UNKNOWN/GENER
Type: 171
Can capture
Can tune
Overlay onto frame buffer
Can clip
Scalable
Video Channels: 4
Audio Channels: 1
Maximum Width: 640
Maximum Height: 576
Minimum Width: 48
Minimum Height: 32
Window Attributes
X Offset: 0
Y Offset: 0
Width: 320
Height: 240
Picture Attributes
Palette: 3 - 565 16 bit RGB
Colour Depth: 16
Brightness: 32768
Hue: 32768
Colour :32768
Contrast: 32768
Whiteness: 0
Channel 0 Attributes
Name: Television
Channel: 0
Flags: 3
Channel has a tuner
Channel has audio
Type: 1 - TV
Format: 3 - AUTO
Channel 1 Attributes
Name: Composite1
Channel: 1
Flags: 2
Channel has audio
Type: 2 - Camera
Format: 3 - AUTO
Channel 2 Attributes
Name: S-Video
Channel: 2
Flags: 2
Channel has audio
Type: 2 - Camera
Format: 3 - AUTO
Channel 3 Attributes
Name: Composite3
Channel: 3
Flags: 2
Channel has audio
Type: 2 - Camera
Format: 3 - AUTO

that is a lot better looking than it was with the all in wonder card in. But that is also because I am rather new to Video in Linux, and getting this thing to work right is a rather pain.

I am pretty sure now that some of the issues is due to that device as unknown/gener because, ever since I got it tweaked and the RPM installed (as opposed to an older source code build) I started seeing these in the syslog:

Nov 21 18:42:30 s_kernel@minazuki.tsukinokage.net kernel: bttv0: timeout: drop=0 irq=21078/21091, risc=3507387c, bits: HSYNC
Nov 21 18:42:34 s_kernel@minazuki.tsukinokage.net kernel: bttv0: reset, reinitialize

so... now... with that in mind... anyone have good advice, besides buying a new card (on the books anyways to EXPAND the CCTV system, but this is needed for testing before I dump more money into the box.) I like the look and FEEL of Zoneminder, all in all, but damn, it can be a pain to get silly things like those error codes sorted.
All in all I just would like someone more in the know to give me a hand with this silly stuff

jameswilson · Post by **jameswilson** » Thu Nov 22, 2007 1:51 am

If your new stop trying to build your own and use the live cd.

Your unknown gereric thing is because linuc cannot detect what card you have and this you need to find out to tell the driver how to talk to it. Most single chip cards work with a card type setting of 77 but it does look like your there but now you need to tell the driver what card you have.
This is done by modifying the modprobe.conf file along thses lines

options bttv card=77

for example.

But the issue you have isnt zm's fault and if your new please try the live cd its so much easier for us too

nombrandue · Post by **nombrandue** » Thu Nov 22, 2007 10:22 pm

Ok, so first thing I have to say is thanks for indirectly calling me a moron. Second thing is the Live CD, while it works, is reporting back the same exact configuration on the system as I had prior, especially with the zmu command.

[root@zm zm]# /union/usr/lib/zm/bin/zmu -d /dev/video0 -q -v
Video Capabilities
Name: BT878 video ( *** UNKNOWN/GENER
Type: 171
Can capture
Can tune
Overlay onto frame buffer
Can clip
Scalable
Video Channels: 4
Audio Channels: 1
Maximum Width: 768
Maximum Height: 480
Minimum Width: 48
Minimum Height: 32
Window Attributes
X Offset: 0
Y Offset: 0
Width: 320
Height: 240
Picture Attributes
Palette: 4 - 24bit RGB
Colour Depth: 24
Brightness: 32768
Hue: 32768
Colour :32768
Contrast: 32768
Whiteness: 0
Channel 0 Attributes
Name: Television
Channel: 0
Flags: 3
Channel has a tuner
Channel has audio
Type: 1 - TV
Format: 1 - NTSC
Channel 1 Attributes
Name: Composite1
Channel: 1
Flags: 2
Channel has audio
Type: 2 - Camera
Format: 1 - NTSC
Channel 2 Attributes
Name: S-Video
Channel: 2
Flags: 2
Channel has audio
Type: 2 - Camera
Format: 1 - NTSC
Channel 3 Attributes
Name: Composite3
Channel: 3
Flags: 2
Channel has audio
Type: 2 - Camera
Format: 1 - NTSC

Since ZM isn't the only product this system is set up to use, in terms of a security system, booting into a custom CD JUST to run the cam software is a bit ridiculous, especially since it yields very little beyond what I already know about the system. So far I think the only differences I have seen with this is the following:

bttv 173020 2 bt878
video_buf 22180 1 bttv
compat_ioctl32 1216 1 bttv
i2c_algo_bit 9032 1 bttv
v4l2_common 7712 1 bttv
btcx_risc 5000 1 bttv
ir_common 9572 1 bttv
tveeprom 14608 1 bttv
i2c_core 21536 3 bttv,i2c_algo_bit,tveeprom
videodev 9408 2 bttv

And as opposed to getting berrated and belittled here again, I will find out what I need to do to get more into this.

While Zoneminder itself is great, I think the experiance I have had here has been what has made my opinion of this soured as much as it is. I am not at all pleased to be treated like crap when I am mostly asking how someone else got it running on a system that is half the power of the one I am working on.

Once again I am confronted with the idea that you have to be an expert in every single chip out there and know how it is supposed to plug into a kernel, and you are supposed to learn this all while not offending the "guru's" who should have enough damned tact not to berate people who are just starting to figure this out themselves. Just because you have learned it before on your own, doesn't mean everyone else needs to be forced to.

>steps off the soapbox<

nombrandue · Post by **nombrandue** » Fri Nov 23, 2007 4:17 am

and for the record, this card uses the same chipset as PV-149 - 8 port video capture card (120FPS) off bluecherry's site (which is said to be linux compatible.

The card chipset is a conexant 878A. I have yet to finish trolling through the halfposts of v4l to get the little chunk of information I need for this card to work.

When I find out, anyone with a RapidOS DVS-8500a, or the Apollo card listed earlier, I will post what I found out, so others who may have this hardware, in terms of capture cards, don't have to dig through google and other search engines, and take a few hundred hours figuring out what the hell has to be done to get the driver in place.

Lee Sharp · Post by **Lee Sharp** » Fri Nov 23, 2007 6:30 am

nombrandue wrote:and for the record, this card uses the same chipset as PV-149 - 8 port video capture card (120FPS) off bluecherry's site (which is said to be linux compatible.

The card chipset is a conexant 878A. I have yet to finish trolling through the halfposts of v4l to get the little chunk of information I need for this card to work.

That is because there is more to a card than the chip. I am amazed that with the linux experience you have you made the classic mistake of buying unsupported hardware. Furthermore you say we are treating you badly, and yet expect instant service. If you want it to "just work" buy hardware on a list of stuff known to "just work." If you want to try some hardware none of us have seen, good luck to you... I can't help with stuff I know nothing about. I did help with the stuff I did know about... But I guess that isn't enough. If you really want to solve some of this, try and eliminate some of the variables so that we can at least work on one problem at a time.