Hi, Running ntopng on a Raspberry Pi 3+ with the latest version for some weeks, I ran a sudo apt-get update and upgrade.Red badges cps-tester.co are often symptoms of problems. For example, if the active hosts ore flows badges are red it means that the ntopng hash tables are full and some ...A good introduction and discussion of how to use timeseries, flows and pcaps to investigate issues is available in blog post Drill Down Deeper: Using ntopng ...

It's been ~24 hours and it seems to still be running. I'll keep an eye on it, but I hope it's stable.

Just seems silly that the package defaults to such a low amount of memory that it can't auto-start. During my googling adventures, I found someone had posted a formula to calculate how much RAM you should allocate, and if that's accurate then I wonder why the devs didn't just compute that for you...

Anyways, I appreciate the response and I'm happy it's now working.

The zeek error above is on pfSense 2.6.0
do anyone have it working on pfSense 2.6.0

@kreki1986 The version number of the Zabbix agent. Look at Package Dependencies shown at the bottom of the package description.

@gertjan said in Heavy traffic monitoring at service or application level:

@jt40

Looks like you could consider using System > Package Manager > Available Packages > ntopng

ntopng builds 'html' pages to show, so no php that parses huge logs files, and goes "time out" doing so.
ntopng can't see what LAN devices 'consume' per process, of course. It can only see : what IP, what ports used.

I'm not using ntopng myself.
Warning : don't install ntopng an walk away. Check disk space used, process power used, etc. daily.

Thanks for the suggestions, but I see that it's not a ready-made solution:

I need to install a couple of things and configure them, I have skill to follow those procedures, but I really don't like to mess up in BSD, I don't know that extremely well :D , same for the distro on top...

It will weight on my system, I have 16GB of Ram and 8 threads, pretty recent CPU, but I don't think it's enough for what I need, not over a couple of months of HA required to catch these network spikes... Eventually, I should set up another machine for that, but it's gonna be another expense...

I've seen that Redis is a DB in memory, I hope it doesn't run in that way in Pfsense as a package, or that at least I can change the configuration.

Is there anything else you would recommend?
I can search online, but it's better to get advices on my specific requirements.

@denis_ju
My static ip was replaced with the hostname "startpage.com".

Otherwise with other hostnames.

@deanfourie Thank you.

Nevermind, I unchecked the option to save settings then reinstalled the package and now I can access the UI.

@luckman212 said in NtopNG and General Latency:

@keyser If I had to guess, I'd say it has something to do with ntop putting all your interfaces into promiscuous mode. I'm sure someone who knows more about the FreeBSD internals could elaborate, but if you're not seeing very high CPU/interrupt usage, then that would be my guess. What kind of NICs are you running?

Yes I agree. But would be nice to have confirmation that this is not a 22.01 specific issue. I can’t recall if I observed such a noticable latency penalty back in the 2.4.x days when testing NtopNG.
Th SG-6100 is running Intel IX and IGC driver based Intel Nics only.

But I see the same on my SG-2100 with its Broadcom MVNETA driver based Nic.

@sameerchouksey do you mean these 2 network segments directly connected to pfsense?

Or do you mean their are either upstream or downstream networks that get to pfsense via a transit network?

@calvinchu hmmm its been a while since I installed and setup ntop, and I only ever turn it on if actually looking for specific data. But I "believe" it defaults to no interfaces selected.. Prob not a good idea if it defaults to "all" interfaces being checked.