Keith via plug on 12 Jan 2021 15:22:16 -0800


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: [PLUG] ZoneMinder / Alternatives


On 1/12/21 3:36 PM, Rich Freeman wrote:
On Tue, Jan 12, 2021 at 3:05 PM Keith via plug
<plug@lists.phillylinux.org> wrote:
On 1/12/21 1:59 PM, Rich Freeman via plug wrote:
Can it alert me in realtime when there is motion in a monitored zone,
both on my phone and on Google Home?  Ie, somebody walks up to the
front door, and a chime sounds throughout the house or something like
that.  It can't happen 20 minutes later.
Yes to all your questions though I can't talk about Google Home. Alerts
to your "phone"... that's a bit ambiguous.  You can get those emailed /
texted to you but they may be other options.
Text would work especially if it can hyperlink to playback the event
or otherwise get me there fast, and if delivery is near-realtime.  I
guess any kind of app monitoring is going to be more
resource-intensive on the phone side, so text might be a good option.
I think there are android front-ends, and if the text message could
integrate with that to quickly pull it up that would be nice.

I'll have to look into sounding chimes/alerts on Google Home.
Yes you get a hyperlink which I really like because its foolproof that way.  As long as I am on my VPN, I can pull the event for that alert quickly.  Near real time I would say is less than 10 seconds depending on how you have it set up.  That's not too bad but what I really want to do is have OH pick up the alarm and then alert me in the mobile app or if I have the OH website / panel up, alert me there.  I know it can be done, I just haven't gotta around too it since the text messaging provides an acceptable amount of time for me to respond to a situation for now but I do want to get it into the 2 to 3 second range.
Can it record both events and continuous footage?  Ie can I have just
a loop of continuous footage that maybe lasts a few days, and then
have events with motion/etc tagged and retained for longer?  That way
if motion is missed I can still review the footage while not letting
storage get too out of hand?
Yes to both and I do both type of events.  You can archive events too so
that they are not deleted but I haven't done that yet.  If you have the
space, I say use it.  YMMV of course, since it depends on you are trying
to do in your environment.  My ZM runs on my LizardFS cluster and right
now is still in a 1Tb container since I haven't need to expand it yet
but I'm sure I will.
Good to hear.  I'll probably also use lizardfs so space really isn't
much of a concern, though bandwidth could be since my nodes are
lower-power.  But video isn't that bad if you're just streaming it and
not seeking a lot.

LFS bandwidth is a thing generally but what I would say is that if you like your LFS performance now then ZM using it should not be an issue.
For footage does it use anything reasonable codec-wise?  I wouldn't
want something super-inefficient like rtjpeg or whatever if I'm going
to record days of full-frame video.

I'm sure I can go ask on one of their forums but I figured I'd start
here and see if anybody is actually using it and what they think...
Typical codecs... ZM isn't doing x265 yet but it does does x264 pass
through which is what I'm using.  Many cams offer MJPEG primary or sub
streams but I don't use them.
Ok, yeah, if ZM is doing its own encoding I could see that getting
CPU-expensive.

I will say this...  I'm running it as a VM and I started out with 2
cores / 4Gb RAM and now I'm up to 8 cores with 8Gb RAM.  I have a 720p
door intercom and 2 4k color night vision cams so far. Long story
short.  ZM need a a good amount of resources.  The busier your system is
the more you would probably need, especially if you are running
continuous, are normally reviewing events or monitoring the system live.
Yeah, that is sounding like more than what I usually do with
containers.  I guess I could just get one camera and play with it and
see how it goes.  I'd be using containers which are more lightweight
than VMs, but if it really needs multiple GB of RAM that is something
I'd have to plan for.
It could be but I don't think so.  Containers are more lightweight but if you need cores, you need cores.  There was a noticeable improvement when I upped the core count.  If I had to make a rule thumb- 2 cores per camera would be a good starting point.  RAM is very much another consideration since rendering video for playback needs it.  Once I added the 4k cams, that became necessary if I recall correctly.
I actually have my old gaming PC sitting around mining ETH right now
since I'm not sure what I want to do with it.  That is a little dated
and suboptimal, but probably plenty for this sort of thing, especially
if it can do NVIDIA GPU encoding.
Remember you're not encoding since you are doing x264 pass-through :) but a gaming PC would be a great place to start.
I would say, compared to the mainstream offerings out there, ZM offers
the same functionality and can do that across a wide array of cameras.
Most security device companies are talking the cloud route and it the
implementations are not going well for a myriad of reasons.
Yeah, for one device the ease of use for the average consumer is going
to be a selling point, but I have these concerns with most of the
options:

1.  Expensive up-front.  They're getting you on the blades AND the
razor it seems.
2.  Ongoing subscription.  Now, cloud-based is actually good for
security since they can't steal your footage, but you're paying a lot
of money for often not much data retention.
Definitely a consideration on the stealing of footage from the local server point of view but there a couple of way to handle that.  One obvious option would be syncing events to offsite storage.  That doesn't specifically require "cloud" and there are straightforward ways to do that.
3.  Limited quality usually - I doubt the cloud services want to be
storing high-quality 4k footage for a month or whatever.  That gets
expensive.
4.  Bandwidth use.  Multiple cameras + cloud either means they're only
storing data in the cloud for events, or your internet connection is
sending a ton of data 24x7.  If those cameras use your WiFi that is
also a significant load.
This is a huge point...  On prem, those 4k streams are only limited by the space I have and when something goes wrong...  I can't even begin to tell you how good 4k looks... and in my case- it looks as good at 2pm as it does at 2am.  There's not that much difference in price difference but mainstream security cameras are definitely doing to be more that the consumer stuff.  Buy once, cry once.
5.  Lock-in.  Those cameras are paperweights if you decide to stop
using that vendor's service, usually.
6.  LImited options.  You might have a few camera options from any of
the consumer-oriented vendors.  With ZM you can mix/match which means
that you can pick the right solution for each location.

There are hybrid solutions like UniFi Protect which stores the data on
your drives and I don't think it has any subscription, but it does
provide remote access via an app/etc.  Those might make sense for
power-users who don't want to deal with full DIY.  However, these
usually don't support stuff like NAS/RAID/etc so storage is going to
be limited and failure-prone.
I started looking at the Unifi stuff too at one point but as much as I like their products the I just abhor vendor lock-in or systems that don't have offerings open enough for people to create the solutions they want.
Thanks for the tips!  That, and the vote of confidence for ZM.

Sure thing.  Feel free to ping me off-list if necessary.

--
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
Keith C. Perry, MS E.E.
Managing Member, DAO Technologies LLC
(O) +1.215.525.4165 x2033
(M) +1.215.432.5167
www.daotechnologies.com

___________________________________________________________________________
Philadelphia Linux Users Group         --        http://www.phillylinux.org
Announcements - http://lists.phillylinux.org/mailman/listinfo/plug-announce
General Discussion  --   http://lists.phillylinux.org/mailman/listinfo/plug