Multi CPU Utilization Graph

[macan]

Anyone ever do this for 24 cores?
2 processors 12 cores each I have snmp data for 24 processors.
Template to share?

Hi,
I prepared small template for multicore CPUs. It displays only avg load for all cores.
http://forums.cacti.net/viewtopic.php?f=12&t=52122
second page

[kakatoos]

hi, i have a problem with the multi cpu.
it shows only 1 random core on the machines i add the template to.
it detects all the cores fine, but shows data only for one.

i use now the ucd/net graph but i have problem with threshold for it, it gives me unreasonable amounts of cpu %

CPU Usage - User [cpu_user] [cpu_user] went above threshold of 80 with 16744869.2667

therefor i try the multi cpu as it should be more accurate, but like you see i have problems with that too.

P.S. any idea why cacti having those gaps in the graph?

[adl0921]

Need 24 multi-core templete, please.

Anyone ever do this for 24 cores?
2 processors 12 cores each I have snmp data for 24 processors.
Template to share?

[xq1]

I am trying to use this template - in imported okay but Some how I deleted my CPU0 CPU1 CPU2 CPU3 datasources so when I create a graph I can not select the data source for each of the CPUs :-~

I have uninstalled and deleted cacti and dropped and readded the Cacti database.

Running on CentOS 6.5

I have been chasing my tail so any advice would be useful

[colifa]

It seems the host MIB tracks threads, and with newer processors I found that 8 was not enough.
Here are some templates built from those in this post for 12, 16 and 20 cores. Hope you find them useful, and please report any errors or issues.

Thank you,

Pablo

[kirush]

What I am doing wrong?
Where I can find data for each CPU?

[Chibiko]

Here for 0.8.8 (2,4,8,12,16,24,32):

[r0bst3r]

Anyone try using report it on these graphs for multiple hosts? The problem i've run into is that report it can report on the individual datasources, but not the totals graphed, as they are not a datasource

[JJX]

Anyone have a template for 64 or even more CPUs?

[MrJackson]

I have updated/corrected Chibiko's graphs and used a new CDEF for all, all of the graphs were using 8 cpu's for the mean calculation and there were a bunch of MIN: items using the MAX: data point.

[JJX]

For an unkown reason the data sources are not saved for the 32 CPU graph.
After pressing "save" all data sources are randomized
I have tested the same scenario using Firefox, IE and Chrome

Before saving:

before_save.png (51 KiB) Viewed 11065 times

After saving:

after_save.png (59.76 KiB) Viewed 11065 times

[JJX]

Anyone tested the latest template?

[shadv]

I am trying to use this on my servers, however the graph never loads. What steps are required to make these work?

[shadv]

Figured it out, I missed the part where you needed to add the SNMP - Get Processor Information under Associated Data Queries.

[rfluharty]

First, I recognize this is a VERY OLD post, but I'm hitting this issue and have searched everywhere for a followup or a solution nor can I find any templates newer than these. The the 12 CPU template works fine, but the 16 and 32 do not. Most of the cores are correctly using right source data, but about 6-8 will not allow me to change the source. (Well, I can change it in the interface, but it does not keep the changes.) I've poured over logs and can't find any reason for this problem. I hope someone can steer me in the right direction. Much appreciation!
I'm running cacti 1.2.8 on Debian bullseye/sid.

For an unkown reason the data sources are not saved for the 32 CPU graph.
After pressing "save" all data sources are randomized
I have tested the same scenario using Firefox, IE and Chrome

Before saving:
before_save.png

After saving:
after_save.png