#zoneminder

:D

I wonder... would I possibly see a significantly lower load if I switched my cameras to "record" instead of "mocord" ??

yes

or is it to be expected that the difference will be minimal

hmmm

mocord still does analysis. Record should not

I would still be able to hover over the timeline graph and get instant snapshots to get a preview of what happened though, right?

if we are talkign h264 streams then the difference mayb enegligible

yeah, h264.

I'm mostly worried about processing power.

not sure what timeloine would give you. I think the hover would work but the graphi would have no data

well if I hover and slowly scroll it gives me an idea of certain things

if I scroll slowly enough I'll be able to see what time the UPS guy is dropping off my new cameras tomorrow, for example :P

yeah timeline is awesome.. just not sure the effect.

if Ican still have that in record maybe it'd be worthwhile for me to switch. I'm nervous that an i3 will get clobbered by two 3mpx + two 720p cams @ 5fps

try it out. seewhat happens

well, I'll find out tomorrow

don't have two of those four yet

well you can do the record vs mocooord experiments now....

yeah, I switched them over, but the 3mpx cameras are the ones (I would think) that would bring the most "brunt" to the server

those 2 are the ones on a truck

(switched them over meaning switched the two 720s I have over to record)

load dropped from about .7 to .4

fancy

moved the other 2 to record, and I'm down to .1.... wow.

when zm is in record does it not record in 10 min segments? my recordings seem a little here, little there. 3 mins, 9 mins, 5 mins, etc

looks like it realigned itself after the 10pm hour changed over

is the script alias in apache config to match the zm_path in options *exactly* to a T? I have /zm/cgi-bin in apache, and /zm/cgi-bin/nph-zms in ZM options. Troubleshooting some socket_sendto errors...

hihi

pid 51367 (zmc), uid 80: exited on signal 11

mmmmm

it's set to dump cores but I can't find any

also, this camera seems to have broken ARP behaviour

Note: you may want to set MySQL to use innodb_file_per_table. Here is a procedure for Ubuntu that should also work for Debian:

wat?

?

?

nm

i just installed 1.29.0

nm

yeah my cams don't work lol

i've never got zoneminder to get a cam:(

where do I put the authentication details?

for a cam

what's the onvif port?

udp 40300

what is that

udp 6970

what kind of firewall access do i need for rtsp? do i need to allow any inbound? such as these udp 40300 and 6970?

hello friends

Hi, I have a DVR that offers a snapshot via an http call of the form "123.123.123.123/cgi-bin/snapshot.cgi?chn=1&u=administrator&p=abcdef&q=1", when I invoke that URL from a browser, it returns a jpg. When I create a monitor with an HTTP/simple source and the host path is the http call that works in the browser, no image is returned. Wireshark sees the packet stream returned to the browser, but Zoneminder never sees _any_ packets. Any ideas why?

p0g0: that's odd, lots of people use monitors that just return a snapshot.

gnulnx: yes, it is odd. It's also a PITA, now days old.

But again, there are no packets sent from the DVR after ZM opens the monitor, and that is a Big Clue.

so if you run curl for that url on the ZM machine does it woork?

I'll try that. All the browsers are happy with it, and in a related way, the mpeg stream has the same issue- vlc can see it fine, but not ZM (I've not checked with wireshark on that call yet tho)

p0g0: hey are you using ZM on FreeBSD

?

no, it is Debian 8 in a VM

ZM is happy with other cameras, it seems fine.

OK

p0g0: Also try the libcurl monitor method

as well as the others.

interesting, curl replies "empty reply". and wireshark sees no packets... so something that works in the browser only works in the browser...

p0g0: white listed browser agent? :(

Could try sending a custom browser header, but are you sure the dvr doesn't require active X or some such nonsense?

some of that is my PICNIC, I tweaked the call by putting the http string in "", that got me a reply via curl, I need to see if it is a valid image...

gnulnx: well, chrome and firefox both get the jpg under linux, so I doubt there is any activeX there.

ah, right

hey, progress... curl does get an image now. Mebbe I am on the right track here.

Why do I have so many events with a NULL EndTime?

How far back is the StartTime?

Some months old, some yesterday.

I figure this must happen when zma crashes.

hey are you still working on the options/console updates or should I take over?

i'm still working on it. I have the options page just about finished but got busy the last week with $paidwork

why would ZM checking for updates in web UI pose a socket_sendto error?

easiest workaround would be to disable update check?

FYI: the "problem" I had was actually - again - an user error. I was trying to get motion detection working but I had "Mocord" as mode. It should have been "Modect".

Spent two days for googling, reinstalling etc and my colleague solved the issue in 2 minutes today :)

zaim_: I don't have access to my box at the moment to disable update check.

it was just recommended on forums.

and I can't understand why?

how often does that happen? I mean, the update trying?

the help says that the update happens infrequently, like once per week. So, if you are seeing constant issues, then it may because of something else. Does the log specifically state that the problem comes from update?

OT: which one is generally better/smarter, the "turret" or the "ring" style cameras? I mean, is the light different and does it make any real difference?

turret: http://www.networkcameracritic.com/Hikvision/ds...

and ring: http://scene7.samsclub.com/is/image/samsclub/00...

oh... turret is not the term I'm looking for. I mean the lighting only, not the mechanism of the pan/tilt

I've not looked at your URLs, but the ring lamps for SLRs are a preffered lighting if you are trying to eliminate shadows.

yeah.. I see the point if the subject is very close to the camera

zaim_: it's not saying that an update is failing. It's saying socket_sendto in the logs. But it happens almost only when I am viewing the feeds live (montage, etC).

But someone just said to disable updates, and another user on another thread said thanks that fixed it.

I'm just wondering... how on earth is that possible