Will, We have about 15k+ devices on Packetfence and don't' see CPU contention but we are running a 3 node cluster. I think the biggest bottleneck I'm seeing is occasional radius timeouts when talking to our Cisco WLC. We are also still on version 11 but plan to update to latest version this summer.
Aaron On Thu, Jun 20, 2024 at 8:24 AM Will Nygard via PacketFence-users < [email protected]> wrote: > Hi Ludovic, > > > > Thanks for the quick response! I’ve edited the service as you noted, and > the CPU load looks a lot better now. > > > > I did notice that the packetfence-fingerbank-collector service is now > constantly restarting, only surviving for a few seconds. Is this the > intended behaviour with this configuration? > > > > Thanks, > > *Will* > > Systems and Network Administrator, ICT Infrastructure > > *From:* Zammit, Ludovic <[email protected]> > *Sent:* Wednesday, June 19, 2024 11:36 AM > *To:* [email protected] > *Cc:* Will Nygard <[email protected]> > *Subject:* [EXTERNAL] Re: [PacketFence-users] PacketFence system > requirements for a large deployment > > > > Hello Will, > > > > Try that: > > > > You can disable the TCP fingerprinting by doing > > > > # systemctl edit packetfence-fingerbank-collector.service > > > > In the editor that opens, add: > > > > [Service] > > Environment=COLLECTOR_DISABLE_TCP_HANDLER=true > > > > Close the editor, then do: > > > > # systemctl daemon-reload > > # systemctl restart packetfence-fingerbank-collector > > Thanks, > > *Ludovic Zammit* > *Product Support Engineer Principal Lead* > > *Cell:* +1.613.670.8432 > > Akamai Technologies - Inverse > 145 Broadway > Cambridge, MA 02142 > > Connect with Us: > > <https://community.akamai.com/> <http://blogs.akamai.com/> > <https://twitter.com/akamai> <http://www.facebook.com/AkamaiTechnologies> > <http://www.linkedin.com/company/akamai-technologies> > <http://www.youtube.com/user/akamaitechnologies?feature=results_main> > > > > On Jun 19, 2024, at 1:39 PM, Will Nygard via PacketFence-users < > [email protected]> wrote: > > > > Hi all, > > > > I’ve been slowly rolling out PacketFence to my K-12 environment, but I’m > running into some performance issues during peak times, such as breaks > between classes and lunch periods. > > > > During these times, I’ve seen CPU spikes as high as 90% utilization. I > have already increased the core count of my VM from 6 to 8 cores, and while > that has helped a bit, I’m a bit scared as to what I’ll need to scale it to > for a full deployment. Right now, I have ~1000 devices registered, but once > fully rolled out that will be ~14000 clients. > > > > My current VM has 8 cores, 32 gigs of ram allocated to it, on a hybrid > SSD/HDD storage setup on Nutanix. Has anyone else with a similar > environment to mine encountered issues like these, and what you might’ve > done to mitigate this, or would my best bet to keep pumping resources into > this one VM, or even scaling out and clustering? > > > > Thanks! > > *Will* > > Systems and Network Administrator, ICT Infrastructure > > > > _______________________________________________ > PacketFence-users mailing list > [email protected] > > https://urldefense.com/v3/__https://lists.sourceforge.net/lists/listinfo/packetfence-users__;!!GjvTz_vk!Uatt1BE9paUcYPov-RwHP_02hvYOEcFd6vz5PLWo5sWjbhpybrMv8kwN2ndjbQn685gfk1iWB5UkpeVu9NAfYgj3nSRoeMltx-0qFA$ > <https://urldefense.com/v3/__https:/lists.sourceforge.net/lists/listinfo/packetfence-users__;!!GjvTz_vk!Uatt1BE9paUcYPov-RwHP_02hvYOEcFd6vz5PLWo5sWjbhpybrMv8kwN2ndjbQn685gfk1iWB5UkpeVu9NAfYgj3nSRoeMltx-0qFA$> > > > _______________________________________________ > PacketFence-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/packetfence-users >
_______________________________________________ PacketFence-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/packetfence-users
