#zoneminder

Morning/afternoon/evening all

happy timezone.

I've setup a new server, primarily for ZM but also for other bits. So i've setup proxmox and have a VM for zoneminder, i've setup LVM on a larger HDD, and a volume for zoneminder that i've mounted to the container

the question is, how the hell do i get zoneminder to see the % of this mounted volume for use with filters?

did you mount it in the appropriate mount-point?

probably not!

i stuck it in /mnt/data

yeah, well. you need to (at least) link it to the storage location for your events.

Titler: ZoneMinder - Wiki - Using a dedicated Hard Drive

o yea bhuddah i'll change the config in a bit

but i'm looking at "df -h" and obviously its not there

knnniggett++

so wondering what zoneminder uses to determine % usage?

ren0v0: then you didn't mount it at all

uh

if it is mounted it will show up in df

well

its mounted from the host

its an openvz container

did you format it with a filesystem? and the like...

/dev/mapper/vgdata-lvzoneminder on /var/lib/vz/private/101/mnt/data type ext4 (rw,relatime,barrier=1,data=ordered)

/dev/mapper/vgdata-lvzoneminder 99G 60M 94G 1% /var/lib/vz/private/101/mnt/data

this is output from the host

well. the host doesn't really matter to your question

so we care mostly about the client now.

well the client can't see the drive, thought that was kind of the point with virtualisation?

there is no output from fdisk -l on the client

i suggest you think about your concept again.

urgh

any suggestions ?

been playing with this for some hours already

was hard enough to get to this point for someone new to VZ

the best would probably be to consult the #openvz guys... if that channel exists

bhuddah,

if i do "df -h /mnt/data" i see the "drive"

on the client or the host?

client

then it's probably ok.

and without the path?

not without the path no

then it's not ok.

can you pastebin the output?

let me restart the CT and see if something happens :D

bhuddah, ok i got it, i was mounting to "private" which is actually an offline filestore, it needed to be /root/ looks like its working now

cool

so assuming i point the config to this DIR, i shouldn't need to do anything else?

or, does it still need knnniggett's wiki link, and bind-mount ?

no.

it'll work if permissions are sufficient

not if he's got it mounted to /mnt/data

it needs to me mounted to the events folder

s/me/be/

wait so the events/images folders are hardcoded?

just realised there is no config for this path, its just "events"

here is how it looks http://pastebin.com/CbU0HbVW

Titler: [Bash] @zoneminder:/mnt/data$ ls -lash total 40K 4.0K drwxr-xr-x 7 root root - Pastebin.com

knnniggett, so i'll still follow your guide and bind the DIR to the /var/cache/zoneminder ?

ah, crap. yes. you need to somehow link it.

symlink will do too.

(at least for me it works)

any benefit/performance on one or the other?

erm, any reason why this isn't actually configurable in the first place?

Please read the FAQ at the bottom of the link (discusses why symlinking can cause you to lose your events in some cases)

knnniggett, FYI, think there may be typo/confusion on that FAQ regarding "newdrive" and "otherdrive", at least it confused me until i realised its probably a mistake

ren0v0: "otherdrive" is part of the name of the mount point the original author chose to use for the example. "newdrive" refers to the physical, new hard drive or volume.

but I can see how that could be confusing

s/newdrive/new drive

fixed

bhuddah: historically the simplest approach to a dedicated volume was to just symlink the volume to the events folder. I used to do it myself. However, now that we have started pushing everyone towards packages (rather than build from source), we found that some packages would not upgrade properly if the events folder was modified. This causes the end user to lose all their events when they tried to start zm after a package upgrade.

that's not nice. i didn't know that.

i would consider that a bug in the package.

I wouldn't. You are modifying the folder structure, which can break any package.

knnniggett, he used "mkdir /newdrive/events"

its a dir not just naming?

just want to be sure we're on the same page before i bork something :)

a loop mount isn't any better.

can i ask again why the path isn't coded as a variable and a default? i'm not trying to sound rude, would help if i could, just curious

would surely get rid of any of this issue

probably "historical reasons"

could be a simple search/replace

doubt it, or it would have been done right? :P

you can submit a pull request and change it.

"would help if i could"

i've not even looked at the code to be fair, but i'm limited to what i can do

i dont even know what its written in, i'll take a peak in a bit with a cuppa

mostly php and c

then no, probably i can't help

though, not through wanting

ren0v0, i keep the default install

but i replace the events dir with a symlink

lrwxrwxrwx 1 root root 15 Aug 4 21:16 events -> /zmstore/events

and /zmstore is actually an nfs mount

actually zm runs on a qemu/kvm VM

hence why using off loaded storage

jzaw, according to knnniggett that isn't the "best" thing to do, so i've followed that wiki page, will see if it works

jzaw, seems we have a very similar setup :)

works :)

stupid question though, its still writing to /var/cache/zoneminder as well ? or is that just a link being shown

knnniggett, you closed this issue, but i still have this warning and a similar ERR https://github.com/ZoneMinder/ZoneMinder/issues... >> Shared data not initialised by capture daemon for monitor SideHigh

just missed an event

morning

when you configure apache to restrict number of parallel handlers for requests, does it not queue additional requests if its reached max limit? An interesting bug was reported by one user - in one of my screens of my app, I am invoking 4 parallel API requests of which 2 were returning HTTP error. The person debugged and noticed that the requests were not reaching syslog and I assumed the problem was with my app. He

later debugged more and found out apache was simply dropping them because he had his max limit set to 3.

[13:11:02] <ren0v0> jzaw, according to knnniggett that isn't the "best" thing to do, so i've followed that wiki page, will see if it works