#Untangle firewall performance reddit Pc
#Untangle firewall performance reddit license

I can get a 46,000+ CPU benchmark PC cheaper than the XG-1541 that is 9000+ CPU benchmark. I want to run on bare metal, on third party hardware, to fulfil the gap that netgate have with their hardware. I do not like the idea of virtualising firewalls either (adds complexity). I looked at buying an XG-1541, but its a 6 year old CPU, its noisy because its rack mounted, and these reasons are key as to why people go and buy third party hardware. They just donвЂ™t sell hardware that can handle IDS/IPS, PFblocker, and any other additional CPU intensive tasks that make a firewall of today. It seems the pfsense / netgate hardware is designed for basic firewalling / IPsec. With a bypass I see ~0.25ms and without it it jumps between ~0.I am actively participating in a discussion in reddit, but IвЂ™d like to have a similar discussion here. It also affects latency, not in a sensitive way, but it's measurable. The moment packets are 'processed' and not just routed, it takes quite a toll on performance (but I don't see heavy CPU usage, maybe some apps can't be multithreaded?). ~4.5Gbps most of the time, even with IPS off. I thought maybe it was one specific app that affects the performance heavily, but I stopped each by each and the results was always the same. I'm considering dropping the apps with a bypass rule as you said. Every server that faces the internet is behind a reverse proxy, but in the event of a compromised server, I thought it wouldn't be 'harmful' to have IPS and firewall filtering the inter VLAN communication (I'm just being paranoid here, it's just a home lab.) It's not a bad idea in fact to make a bypass rule from any non-WAN to any non-WAN. The Untangle server has enough CPU to do inter-VLAN traffic at line speed, so I don't mind going that route. I'd try to avoid adding another device for inter-VLAN traffic, I'm already pretty high on the power bill (the L2/е…ѓ Mikrotik switch could do that, but it drops performance to less than 1gbps, it doesn't have enough CPU, that's why it's only switching). I'd like to avoid bypass rules, in fact I'm comfortable with the performance I'm getting but it would be nice to tune it further if it's possible. I don't see significant CPU nor RAM usage.ДЄre there any knobs to tune to try to get closer to line speed, or get better performance out of the server? It's quite capable, I'd say (6 cores at 3Ghz with 16GB RAM). If I disable the bypass rule, I have lots of retries and performance drops significantly between 3 to 5gbps. If I make a 'bypass' rule to allow VLAN10 to VLAN20 to be bypassed, it will be routed by Untangle, and the performance will be line speed: ~9.45Gbps I have the follow apps running on the Untangle server: Firewall, Ad Blocker, Web Filter, Bandwidth Control, Application Control, Intrusion Prevention, Policy Manager, Directory Connector and Reports.

#Untangle firewall performance reddit license

I have Untalge Home (the version before the new license mode), and I a run a home lab with several servers and VLANs.ДЈ ESXi servers (Xeon 2678v3, 256GB RAM, 2x10Gbps)ДЎ TrueNAS server (Xeon 2680v1, 96GB RAM, 2x10Gbps)ДЎ Primary Worksatation (Ryzen 5800X, 64GB RAM, 1x10Gbps)ДЎ Untangle Router (Core i5 8500, 16GB DDR4, 256GB NVMe, 2x10Gbps)Д®verything L2 works at line speed (~9.45gpbs) because it doesn't go up to the Untagle Router. I was wondering if there are any performance tuning options for 10GbE.