01-12-2016 01:27 AM - edited 03-10-2019 06:32 AM
Hi Guys,
I have looked at the link below to try and understand how firepower impacts throghput and have put the following chart together.
I then had a call with Cisco TAC, and he mentioned the bandwidth is halfed when URL service is enabled and further halfed when AMP is enabled.
This is not on cisco website. Can anyone confirm and give an opinion on this?
how the Max throughput is achieved. Is it per connection regardless of load or under maximum load/maximum connections? If per connection, can one assume a 5506 would give a significantly higher throughput than 23Mbps if few users are passing traffic through the ASA and all firepower functionalities are enabled ?
Model\Feature |
Maximum Performance without Firepower(Mbps) |
Maximum AVC Throughput(Mbps) |
Maximum AVC+IPS Throughput(Mbps) |
Maximum AVC + IPS + AMP Throughput(Mbps) |
Maximum AVC + IPS+ AMP + URL Filtering Throughput(Mbps) |
Maximum VPN Throughput(Mpbs) |
5506 |
300 |
250 |
125 |
45 |
23 |
100 |
5508 |
500 |
450 |
250 |
90 |
45 |
175 |
5512 |
500 |
300 |
150 |
50 |
25 |
200 |
5515 |
600 |
500 |
250 |
75 |
38 |
250 |
5516 |
900 |
850 |
450 |
150 |
75 |
250 |
5525 |
1024 |
1100* |
650 |
188 |
94 |
300 |
5545 |
1536 |
1500 |
1000 |
288 |
144 |
400 |
5555 |
2048 |
1750 |
1250 |
363 |
182 |
700 |
01-13-2016 12:39 AM
Throughput is often measured using an IMIX profile.
01-13-2016 05:54 AM
Your exact numbers are off a bit but the overall idea that URL filtering and especially AMP will reduce firewall throughput is correct.
Because there are so many variables that go into a given unit's throughput, Cisco does not publish all of the various scenario numbers publicly. Your Cisco or partner SE can discuss with you in detail considering your use case and size the appliance appropriately based on that.
The testing methodology does generally assume a large number of flows, as per the IMIX profile that Philip mentioned.
07-09-2016 07:03 PM
I assume using an IMIX profile would return throughput less than a speedtest, for example.
We have an ASA5515. Without Firepower, we get about 900 mbps to a decent speedtest site.
We are blocking Torrents, Malware, but no IDS. We are getting 300 mbps down/220 up.
Latency jumps to 200-300ms during the test, from a second client. Jitter is huge, listening to audio and doing a VoIP call seem unaffected during the test.
We are going to have to go to a ASA 5555 I fear. We have two of the ASA 5515-X, so we might try clustering first.
-Robert
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide