Suricata can (probably) only be used as IDS on our routers; snort as IPS.What is the advantage of Snort vs Suricata?
What is the advantage of Snort vs Suricata? Sounds like they both do the same thing.
We've been warned that Snort if very CPU intensive on a 86U. Is that also true for Suricata?
Also, for CPU utilization, how well does Snort play with Unbound and Skynet and Diversion?
This depends on the router. AC86U does have problems running Snort, it needs active cooling while the AX88U can easily handle the additional load from Snort without any major down sides. It does reduce the internet throughput slightly, but so does Suricata.Unfortunately, the former process (IPS) can be so taxing to the processors and memory of our routers that they can't properly do what they're really designed for without hurting themselves in the process - it's a case of the spirit being willing but the body being rather unable.
(VPNs are another layer of security - protect the clean data that passes between protected points - but you have to trust (verify!) that the other point is as (or more!) secure than your network)
I have my snort3 running for days, I had to stop and restart last night to check my settings before posting.. I didn't have any issues with sudden stoppage. Since I don't have AX88U, I would monitor the memory usage to make sure it didn't "run out of memory".Tonight I checked on how Snort and the router was doing. Unfortunately Snort wasn't running any longer and the system log was overwritten at 17:00h and after that there was no error. So it must had happened before that. I'll keep monitoring more closely now. Anyone else having the same problem?
cat /proc/buddyinfo
echo 1 > /proc/sys/vm/compact_memory
correct, and people should be aware of the tradeoff between security and speed. Thankfully in a properly implemented strategy with the correct tools, it is hardly noticeable - and the ones we're discussing here are some of themThis depends on the router. AC86U does have problems running Snort, it needs active cooling while the AX88U can easily handle the additional load from Snort without any major down sides. It does reduce the internet throughput slightly, but so does Suricata.
drop tcp $HOME_NET any -> <an IP address> any (msg:"test";)
wget <same IP address> --no-check-certificate
The rules are loaded in RAM (or cached in RAM), so if you updated the rules, you need to restart snort. Also you need to make sure snort.lua knows where those rules are (if you haven't already pointed to them).I extracted a package with rules downloaded to the folder for rules. are they automatically reloaded or I need to restart snort?
Tried this rule, but several minutes after starting snort I can still access facebook.comdrop tcp any any -> any any ( msg:"Facebook trafic Seen"; appids:"Facebook";sid:10000001; )
afpacket DAQ configured to inline.
Commencing packet processing
++ [0] eth0:br0
Did you load OpenAppID? My example rule was using OpenAppID to help filtering app specific traffic. Also check my 1st post where I posted my reference snort.lua file and see if you enabled OpenAppID correctly. Also use the Validation Cmdline to make sure all the rule sets are loaded properly.Tried this rule, but several minutes after starting snort I can still access facebook.com
Also, before starting snort the AC86u's temperature was 79°C and it remains at that value (does not increase)
This is all I see from snort (except from some rule statistics before that):
Code:afpacket DAQ configured to inline. Commencing packet processing ++ [0] eth0:br0
/opt/var/log/snort/alert_fast.txt remains empty
Is it doing anything at all? (Core 1 and Core 2 fluctuate a lot, but the sum is always around 110%)
Happy to report Snort is running now since Sunday without any issues. Now if I have some time I'll fiddle around with some of the rules and settings.I have my snort3 running for days, I had to stop and restart last night to check my settings before posting.. I didn't have any issues with sudden stoppage. Since I don't have AX88U, I would monitor the memory usage to make sure it didn't "run out of memory".
Background:
Linux kernel has a strange way of managing memory, if the memory manager wanted a memory of certain size (let's say 2 Megabytes) and if the system memory is "fragmented" in a way where you had LOTS of 4K, 8K and 16K memory chunk but you didn't have a memory chunk to meet the 2 megabytes memory demand, it would declare itself "out of memory". With my custom fork, I backported newer memory management enhancements from Linux 4.4 all the way to Linux 5.8 to our older Linux 4.1 to address this funny out of memory situation (this happens even MORE often on an AC86U due to only having 512 MB of RAM) by "defrag" those holes and compact them together to make larger chunks available for other uses.
gives a glimpse into these memory chunks, if you see lots of 0's in between it means for those chunk sizes aren't available and if an app ask for those chunk sizes, your system could be out of memory and may have unexpected behavior.Code:cat /proc/buddyinfo
you can doto alleviate this memory fragmentation but with stock Linux 4.1 kernel, that memory defrag can't compact as well as the newer releases (or my hybrid kernel).Code:echo 1 > /proc/sys/vm/compact_memory
I'm sorta sem-retired from Android scene for a few years now (I check in every once in a while to see what new cool things are happening with Android), with just a bit of spare time remaining (work + family), I hack a few things I use on a daily basis (such as AsusWRT and some small open source projects).Wow @faux123, youre not only doing android kernel stuff but also stuff for AsusWRT? Nice
Welcome To SNBForums
SNBForums is a community for anyone who wants to learn about or discuss the latest in wireless routers, network storage and the ins and outs of building and maintaining a small network.
If you'd like to post a question, simply register and have at it!
While you're at it, please check out SmallNetBuilder for product reviews and our famous Router Charts, Ranker and plenty more!