#zoneminder

ren0v0_: did you ever try an actual vm?

shitdamn....snapshotted a live system and fubar'd it...yay

cool....reinstalled zm and back up and running

all in all took me about 20 minutes to get it up and running like it was again

xnxs, yes i have KVM setup but getting 100s of ERR llogs

all quite similar to the other container

i have no idea whats going on...i just did a reinstall and absolutely no issues

meh

ren0v0_: Yes, like i said, the container is unable to clear the memory cache, only the host (kernel) can, and it does it when the memory runs too low. If you will watch free -m, you will see that your free memory gets to zero, and only then, it frees the cache

ren0v0_: I once read about a workaround for this, but never had the time to test it. The workaround is fairly simple. Instead of setting the soft&hard limit identical (or not setting the soft limit at all), set the soft limit to like 80%

Currently as far as i remember, Promoxmox sets them both the same, while LXD sets the soft limit to 90%

mastertheknife, where are these limits set?

actually i was referring to my new KVM install that has all the errors :D

ren0v0_: LXC uses cgroups for these limits

the shared memory error is weird, saeems like a bug

not a UI option then ?

the UI option sets them both to 100%

SHM stops at 24%, regardless of how many montitors i start

currently I only have 3 recording and i'm getting this Got empty memory map file size 0, is the zmc process for this monitor running?

and what do you see in the zmc log

07/04/18 08:38:27.996690 zmc_m3[1744].WAR-zm_monitor.cpp/2390 [Return from Capture (-1), signal loss]

07/04/18 08:38:27.997706 zmc_m3[1744].ERR-zmc.cpp/309 [Failed to capture image from monitor 3 (1/1)]

trying to enable debugging logs now

master here is something https://pastebin.com/iUeRWNxw

and this

07/04/18 08:45:16.386159 zmc_m3[2419].DB1-zm_monitor.cpp/557 [Unable to map file /dev/shm/zm.mmap.3 (829446960 bytes) to locked memory, trying unlocked]

07/04/18 08:45:16.386174 zmc_m3[2419].DB1-zm_monitor.cpp/561 [Mapped file /dev/shm/zm.mmap.3 (829446960 bytes) to unlocked memory]

same stuff over and over

yeah....i have no idea wtf is going on....

i just did a new install on a kvm on proxmox and had zero issues with 10 monitors on it

xnxs, must say i'm using 1.31.44 so completely different

but it seems this may be related to a specific monitor or two for me

this one specifically is WiFi

mastertheknife, does that sound about right, from what those logs show? Is there anyway i can force ZM to be more fault tolerant or something, rather than crashing everytime

nope...i just installed 131.44 5oo

too

the KVM was far far worse for me

10000s of errors there

so i've gone back to this LXC one

i even went to ubuntu 18.04 this time too

07/04/18 08:55:27.270040 zmc_m3[2470].INF-zm_ffmpeg_camera.cpp/707 [Unable to read packet from stream 0: error -541478725 "End of file".]

this is where its failing every time

grab ffmpeg and dump the camera to a file on the cli to see if it shits itself?

yea i'm looking into that as we speak actually

i don't use wifi for any of my cams....

Have to for 2 of them currently

run a cable to them?

i tested ffmpeg on a wired one and still got this

[NULL @ 0x10592e0] RTP: missed 1 packets=00:07:48.03 bitrate=N/A

Last message repeated 3 times

not that easy atm :D

:P

keeps reporting wrong FPS also

Capturing at 333.33 fps,

are you specifying a fps in the settings?

i dont no

but then i see in latest version ic0n added 30 here, so i put 30 for them all

but this was the case for a long time that it reports an extremely high fps when initialising i think

I've stopped some other monitors, SHM is not at 14% and it seems a little more stable

ffmpeg running no problem

this is so annoying

wondering if its just networking now

i'm recording 10minute events, so far on this camera i dont have a single one that has made it to 10minutes

the other wifi camera has made 2 now

trying to see if when it crashes in ZM what ffmpeg does

lol not crashing not

now *

the only thing i've done is turn off some monitors, so let me put them back on

this is ridiculous lol, started another 5 monitors and not a single crash yet

mastertheknife, where was that setting in proxmox to change hard/soft limits

as i'm back to 20GB in buffer

its weird, now that the buffer is full SHM stops at 37%

added more and more monitors, increased FPS to 20 and no change there

the only thing i can see is that the wifi access point, channel gets highly utilised at some points, maybe that could impact it

but its on the best channel, so nothing else i can do tehre

ok caught one

[NULL @ 0x1ec8340] RTP: missed 1463 packets:04:05.87 bitrate=N/A

[NULL @ 0x1ec8340] sps_id 5 out of range

Here is the matching paste from ZM, if anyone can spot something that would be great https://pastebin.com/8YrMfNYR

mastertheknife, are you sure about this being LXC and not ZM regarding buffers? because when i stop ZM it frees up

with zoneminder stopped the cache still sits there, could this be mysql ?

anybody here running zoneminder on nginx ? I need help with the config. I get a 404 when trying to access http://ip/zm

void9, what version ?

compiled from git master, nginx 1.14

zoneminder git, nginx 1.14*

no zoneminder

just fixed it with the help of a friend, i had two server blocks on the same port :D i thought it's allowed if they dont point to the same location

mastertheknife, I've spoken with some proxmox guys for a long while, and whilst you're right that LXC may not be able to clear these caches, why the hell are they filling up so quickly? what exactly is using them? I've got 26GB of RAM on this container now, and after 30minutes there is only 1GB free, the rest in cache

I can force clear caches from host, but what is zoneminder doing?

Also, why is this even happening if SHM is only at 14% ??

anyone try the git version? does the web interface render properly ?

looks pretty bad here, like no skin

change CSS to "base" in options

void9, ^

as a matter of fact I just did that, still looks the same

and the add button for new camera doesn't do anything :\

screenshot

my guess is your apache config isn't right

which is why i asked what version of zoneminder you are talking about

here it is: https://imgur.com/a/Zb6wZUJ

and it's nginx

go to "display"

there is another skin option there

this is 1.31.44 ?

can just make it out top right

yes

i have no idea if someone has made an nginx config for this, i would suggest changing to apache and using iconnor's apache config.

please refer to this thread https://forums.zoneminder.com/viewtopic.php?t=2...

:( i went through some effort to use nginx :D

I would suggest this https://gist.github.com/connortechnology/9ca83f...

why?

There is no issue using nginx, you just need to know how to use it, and translate the required settings to nginx :)

Here is what the packaging scripts need to do:

- ZoneMinder uses a cache (busting) folder now. This folder path needs to be set at build time using -DCMAKE_CACHEDIR

- Packaging scripts need to ensure the new cache folder is owned by the web server account. The name of this account is different for every distro

- The apache config file that ships with the package has to be updated to incorporate the new features. A generic template can be found here: https://github.com/ZoneMinder/zoneminde ... he.conf.in

should I change this compile option ? -DZM_CONTENTDIR=/var/cache/zoneminder \

and what should I set -DCMAKE_CACHEDIR to ?

in another place I've seen: There is now a web cache folder. You need to specify its location during build with the cmake flag -DZM_CACHEDIR

confused :/

recompiled with that option.. removed cache, and layout is fixed :)

ok now, any tutorials for getting the streaming links from chinese ip cameras ?

there are, but theyre in chinese

the rtsp port is open but i can't get it to stream, in zoneminder or vlc

I use hikvision