You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
SniffNet Remote Agent to Monitor the Network Adapter on a Remote Host or Hosts #303
Setup a Remote Agent for SniffNet, just like Nagios's Remote Agent to Monitor the Network Adapter on a Remote Host or Hosts.
This would be a Option just like in SniffNet when you Pick which Network Adapter you would like to Monitor Network Traffic.
The Remote Agent would be Installed on Windows/Linux and Send Live Data of Network Traffic Information back to SniffNet.
Is your feature request related to a problem?
No response
Reacted by hbednar, reignmaker, Vlad Bokov, Giuliano Bellini, T-256, AndreKR, Kevin Cali, FDrebin and Richard Walmsley
This is now implemented thanks to PR #1270
The feature is now on the main branch, and will be officially released in a couple months with Sniffnet v1.6.0.
As of now, the available adapters are those identified by pcap and are only internal to the machine on which Sniffnet is installed.
The idea you mentioned would require the development of the remote agent, if I understood correctly, and this is out of scope at the moment.
If you are aware, feel free to link some resource from which I can take inspiration for the modalities of work of remote agents and the transmission of data to the main application.
If the process wouldn't break too much the existing code base, I could consider this feature in the long term.
Thanks!
This seems a bit out of scope at the moment, since many other features are being worked on and none of them is going in the direction of a server + agent application, at least for the moment.
Feel free to leave this issue open, but I'll tag it as wontfix for now.
I don't think any of these solutions will provide anything more than RX TX counters for each interface. Nagios plugins aren't exporting port and IP src/dst addressing for each flow, it's just a count of how many packets came into and out of each individual network interface.
@GyulyVGC@kenrmayfield
Opensnitch might be a good option, its a firewall made of a daemon and a UI.
But, it can run the daemon without the UI on a remote clients and send the logs to the UI on a different server.
It also supports SIEM integration, so either way you don't need to maintain an agent just read the logs.
This seems a bit out of scope at the moment, since many other features are being worked on and none of them is going in the direction of a server + agent application, at least for the moment.
Feel free to leave this issue open, but I'll tag it as wontfix for now.
I included this request in the project's roadmap, but it'll be likely one of the last features to be implemented.
I'm seriously considering to introduce this feature in one of the next major releases of the app, probably version 1.6 or 1.7 (see the new Roadmap).
Starting to think about the implementation, I was exploring different possibilities for the protocol to use for data transmission between agent and Sniffnet.
A possible strategy that comes to my mind is to use gRPC, but I'm open to your suggestions.
Reacted by SeimusS, AndreKR, FDrebin and Richard Walmsley
@GyulyVGC
Having an agent would be excellent but, it may be a lot of extra work and time to maintain it.
Have you thought about using an existing tool and just collecting the data from that. Then you wouldn't be responsible for maintaining an agent but ownley working with its data.
I am not sure if it would work but as an example there is opensnitch and it allows sending data to SIEM
Using a Existing Tool for Collecting Data is not a good idea for SniffNet. If a User wants to Collect Data from Third Party then they would use something like Grafrana, InFluxDB and Telegraf.
You want your on STAMP on SnifNet without Third Party.
A Agent specifically for SniffNet is the Best Way to Go.
Yeah, I'm in fact of the idea of developing a dedicated Sniffnet agent.
While I understand that it'd introduce some maintenance efforts, I plan to keep it as a simple packet capture agent with limited configurations.
I just have to determine which is the best way to transmit data (e.g., gRPC, syslog, raw sockets, etc...).
By "best" I mean that we should satisfy the following:
Hey everyone, I plan to start with the development of this feature in the coming weeks so this is one of the last chances to express feedbacks.
What I plan to develop is the following:
a Sniffnet agent exporting network flows in IPFIX format
new capabilities for the Sniffnet app itself to be an IPFIX collector, allowing to gather and show data from multiple IPFIX exporters simultaneously (either from Sniffnet agents or other IPFIX producers)
I often get requests about including support for OPNsense and OpenWRT, and I think such requests could be satisfied by this one.
Please correct me if I'm wrong and in such case feel free to share more about what you'd expect.
I don't know how the architecture will work yet, but having agent support for headerless servers would be great.
Maybe by having a web UI for headerless servers or sending logs to a device with a desktop UI, like OpenSnitch Firewall does.
I just had an insightful conversation with the creator of NetGauze Ahmed Elhassany, and IPFIX seems to actually be suitable for what Sniffnet needs to implement this feature, so I'll move forward with the IPFIX-based implementation.
This is now implemented thanks to PR #1270
The feature is now on the main branch, and will be officially released in a couple months with Sniffnet v1.6.0.
Here is a preview:
ipfix-multi-exporters.mp4
Reacted by zero77
Reacted by Richard Walmsley, Benjamin Kane, AndreKR and T-256
Activity
GyulyVGC commentedon Jul 6, 2023
As of now, the available adapters are those identified by pcap and are only internal to the machine on which Sniffnet is installed.
The idea you mentioned would require the development of the remote agent, if I understood correctly, and this is out of scope at the moment.
If you are aware, feel free to link some resource from which I can take inspiration for the modalities of work of remote agents and the transmission of data to the main application.
If the process wouldn't break too much the existing code base, I could consider this feature in the long term.
kenrmayfield commentedon Jul 8, 2023
GyulyVGC commentedon Jul 8, 2023
Thanks!
This seems a bit out of scope at the moment, since many other features are being worked on and none of them is going in the direction of a server + agent application, at least for the moment.
Feel free to leave this issue open, but I'll tag it as
wontfixfor now.aderusha commentedon Jul 15, 2023
I don't think any of these solutions will provide anything more than RX TX counters for each interface. Nagios plugins aren't exporting port and IP src/dst addressing for each flow, it's just a count of how many packets came into and out of each individual network interface.
hbednar commentedon Aug 2, 2023
@GyulyVGC @kenrmayfield
Opensnitch might be a good option, its a firewall made of a daemon and a UI.
But, it can run the daemon without the UI on a remote clients and send the logs to the UI on a different server.
It also supports SIEM integration, so either way you don't need to maintain an agent just read the logs.
https://github.com/evilsocket/opensnitch/wiki/SIEM-integration
kenrmayfield commentedon Sep 2, 2023
@GyulyVGC
Have you thought about making a Agent for Windows and Daemon for Linux so that SniffNet can Sniff
Remote Machines?
GyulyVGC commentedon Sep 2, 2023
Hi @kenrmayfield, as I anticipated:
I included this request in the project's roadmap, but it'll be likely one of the last features to be implemented.
GyulyVGC commentedon Jun 19, 2025
I'm seriously considering to introduce this feature in one of the next major releases of the app, probably version 1.6 or 1.7 (see the new Roadmap).
Starting to think about the implementation, I was exploring different possibilities for the protocol to use for data transmission between agent and Sniffnet.
A possible strategy that comes to my mind is to use gRPC, but I'm open to your suggestions.
zero77 commentedon Jun 19, 2025
@GyulyVGC
Having an agent would be excellent but, it may be a lot of extra work and time to maintain it.
Have you thought about using an existing tool and just collecting the data from that. Then you wouldn't be responsible for maintaining an agent but ownley working with its data.
I am not sure if it would work but as an example there is opensnitch and it allows sending data to SIEM
https://github.com/evilsocket/opensnitch/wiki/SIEM-integration
kenrmayfield commentedon Jun 19, 2025
@GyulyVGC
Using a Existing Tool for Collecting Data is not a good idea for SniffNet. If a User wants to Collect Data from Third Party then they would use something like Grafrana, InFluxDB and Telegraf.
You want your on STAMP on SnifNet without Third Party.
A Agent specifically for SniffNet is the Best Way to Go.
GyulyVGC commentedon Jun 19, 2025
Yeah, I'm in fact of the idea of developing a dedicated Sniffnet agent.
While I understand that it'd introduce some maintenance efforts, I plan to keep it as a simple packet capture agent with limited configurations.
I just have to determine which is the best way to transmit data (e.g., gRPC, syslog, raw sockets, etc...).
By "best" I mean that we should satisfy the following:
zero77 commentedon Jun 19, 2025
if you want to develop your own agent, then i would go with your original comment of using gRPC
6 remaining items
GyulyVGC commentedon Apr 20, 2026
Hey everyone, I plan to start with the development of this feature in the coming weeks so this is one of the last chances to express feedbacks.
What I plan to develop is the following:
I often get requests about including support for OPNsense and OpenWRT, and I think such requests could be satisfied by this one.
Please correct me if I'm wrong and in such case feel free to share more about what you'd expect.
zero77 commentedon Apr 21, 2026
I don't know how the architecture will work yet, but having agent support for headerless servers would be great.
Maybe by having a web UI for headerless servers or sending logs to a device with a desktop UI, like OpenSnitch Firewall does.
kenrmayfield commentedon Apr 21, 2026
@GyulyVGC
How you been?
Excellent News to hear.
This is going to be a Great Edition to SniffNet.
I think this is going to push SniffNet even more over the top.
GyulyVGC commentedon May 11, 2026
I just had an insightful conversation with the creator of NetGauze Ahmed Elhassany, and IPFIX seems to actually be suitable for what Sniffnet needs to implement this feature, so I'll move forward with the IPFIX-based implementation.
GyulyVGC commentedon May 11, 2026
@all-contributors please add @ahassany for ideas.
allcontributors commentedon May 11, 2026
@GyulyVGC
I've put up a pull request to add @ahassany! 🎉
kenrmayfield commentedon May 13, 2026
@GyulyVGC
Great News!!!!
GyulyVGC commentedon Jun 12, 2026
A first step to finally close this issue is complete: I present you
sniffnet-agent, the unit in charge to export traffic flows in IPFIX format.Now what's missing is only the support for collecting IPFIX at Sniffnet-side, which I already started to work on in the
ipfixbranch.kenrmayfield commentedon Jun 19, 2026
@GyulyVGC
How you been?
Good News about the SniffNet Agent.
GyulyVGC commentedon Aug 25, 2026
And... boom! 💥
This is now implemented thanks to PR #1270
The feature is now on the
mainbranch, and will be officially released in a couple months with Sniffnet v1.6.0.Here is a preview:
ipfix-multi-exporters.mp4