12-27-2024 12:30 PM
I have an architecture with two 3110 chassis, each with 3 instances. These instances are in active/standby HA, and when I have all three instances on the same chassis, the memory of Snort 3 increases, and we start having issues with the applications. The sizing of the instances individually looks fine. Does anyone have any suggestions or experience with what might be happening?
12-27-2024 03:27 PM
what version you are running and can you provide more details ?
what is the memory usage over time from the time you start the box and over time ?
does this happen right away or over time ?
Typically there is more contention for CPU than memory. But you may be run into defects like a memory leak.
what are you traffic rates through each box ?
are you offloading some trusted traffic using pre-filter without using IPS/SNORT, that may alleviate the load on snort.
But, memory issues etc need multiple iterations of troubleshooting and diagnostic commands, and if you have support, i would contact cisco TAC to take a look.
12-29-2024 08:21 AM
version 7.4.2
It happens over time
12-29-2024 10:26 AM
If it happens over time, most likely a leak (bug) that is increasing over time. Packets etc are transient and only hold memory during the transmission and should be released. 7.4.2 is the latest, so you may have to open a TAC CASE to troubleshoot this further.
12-28-2024 02:01 AM
What Idea of using three instances ?
MHM
12-29-2024 08:19 AM
we are separating in the instances, edge, dmz, users
12-29-2024 11:34 AM
why I ask?
You need to management resource it not need for traffic pass from internal to DMZ then via VPN to pasd two deep inspect' one in internal and other in DMZ' same sa traffic pass from internal to internet it not need to deep inspect by edge and internal.
MHM
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