08-12-2022 06:15 AM - edited 08-12-2022 06:17 AM
Hi, we have two Nexus C93180YC-FX3 working in vPC and they are constantly between 60-70% of memory usage.
show system resources
Load average: 1 minute: 0.72 5 minutes: 0.80 15 minutes: 0.87
Processes : 690 total, 1 running
CPU states : 3.77% user, 2.77% kernel, 93.45% idle
CPU0 states : 1.00% user, 1.00% kernel, 98.00% idle
CPU1 states : 2.02% user, 2.02% kernel, 95.95% idle
CPU2 states : 2.04% user, 2.04% kernel, 95.91% idle
CPU3 states : 2.97% user, 0.99% kernel, 96.03% idle
CPU4 states : 4.00% user, 2.00% kernel, 94.00% idle
CPU5 states : 0.00% user, 9.78% kernel, 90.21% idle
CPU6 states : 1.01% user, 2.02% kernel, 96.96% idle
CPU7 states : 15.68% user, 1.96% kernel, 82.35% idle
Memory usage: 16332592K total, 9827480K used, 6505112K free
Kernel vmalloc: 0K total, 0K free
Kernel buffers: 391356K Used
Kernel cached : 4567440K Used
Current memory status: OK
The top processes which are consuming more memory:
3140886528 l2fm
2426183680 tahusd
1301229568 swtele
1225977856 libvirtd
1039392768 netstack
1004556288 m6rib
I am trying to find which memory values are considered as "normal".
Thanks.
08-12-2022 06:37 AM
""
08-12-2022 11:07 AM
There is hard to pinpoint to a "normal" memory value, because networks are not the same and switches are used differently by each customer. This means your "normal" might not be someone else's "normal".
However, what you can do is monitor these values (snmp, telemetry etc), and build your baseline. This way, you can see what is your normal and if there are any deviation from it. More importantly to memory usage, monitor if there is constant increase over time, with no changes in the network. If yes, then you should investigate it further.
Stay safe,
Sergiu
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