cacti not listing cpus

[tbsh]

Hi,

I'm seeing a weird problem in cacti, we have a cacti server running that is monitoring multiple hosts and only one host is affected by this problem.

The problem is, for this host we can't graph any of the CPU Utilization. I've done some debugging but this is where I'm stuck.

From the cacti interface when I click on (Verbose Query) next to SNMP - Get Processor Information I get the following output.

Code: Select all

+ Running data query [9].
+ Found type = '6 '[script query].
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'
+ XML file parsed ok.
+ Executing script for list of indexes '/usr/local/bin/php -q /usr/local/share/cacti/scripts/ss_host_cpu.php 192.168.1.2 11 2:161:500:1:10:public:::MD5::DES: index'
+ Executing script query '/usr/local/bin/php -q /usr/local/share/cacti/scripts/ss_host_cpu.php 192.168.1.2 11 2:161:500:1:10:public:::MD5::DES: query index'
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'

As you can see there is no CPU's listed but running this same command in a shell I get the CPU's listed as below

Code: Select all

:~ $ /usr/local/bin/php -q /usr/local/share/cacti/scripts/ss_host_cpu.php 192.168.1.2 11 2:161:500:1:10:public:::MD5::DES: index
0
1
2
3

As I've mentioned this is only affecting 1 host out of 23
Any help will be appreciated.

Thanks

[Hyperlord]

Hi,

different snmp.conf on that machine?

[tbsh]

Hi

There are some changes but that's only to listen to different services like mailq on one and apache on another but the rest are all the same

[tbsh]

Hi,

Can someone please help with this problem, any suggestions will be helpful.

Cheers

[gandalf]

Is the community shown exactly as it is? We knew, that strange characters broke things in the past
R.

[tbsh]

Hi

Yes the community is the same for all hosts, its different from the post for security reasons and its all alphabetical characters, no numbers or symbols.
They are a mixture of capitals and lower case characters though.

[n1ck3l]

When this happens, you can check it the Version of the agent. If it's in linux and using net-snmp , try the command "sudo snmpd -v". Check if this is newer than 5.4.X.Y; if so, it may have changed the way the agent net-snmp reports the cpu per user..

You can try the data query I just uploaded (check my posts), and see if you got rows from it; then you can use the data sources to fill a graph.

Good Luck!

Hi,

I'm seeing a weird problem in cacti, we have a cacti server running that is monitoring multiple hosts and only one host is affected by this problem.

The problem is, for this host we can't graph any of the CPU Utilization. I've done some debugging but this is where I'm stuck.

From the cacti interface when I click on (Verbose Query) next to SNMP - Get Processor Information I get the following output.

Code: Select all

+ Running data query [9].
+ Found type = '6 '[script query].
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'
+ XML file parsed ok.
+ Executing script for list of indexes '/usr/local/bin/php -q /usr/local/share/cacti/scripts/ss_host_cpu.php 192.168.1.2 11 2:161:500:1:10:public:::MD5::DES: index'
+ Executing script query '/usr/local/bin/php -q /usr/local/share/cacti/scripts/ss_host_cpu.php 192.168.1.2 11 2:161:500:1:10:public:::MD5::DES: query index'
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'
+ Found data query XML file at '/usr/local/share/cacti/resource/script_server/host_cpu.xml'

As you can see there is no CPU's listed but running this same command in a shell I get the CPU's listed as below

Code: Select all

:~ $ /usr/local/bin/php -q /usr/local/share/cacti/scripts/ss_host_cpu.php 192.168.1.2 11 2:161:500:1:10:public:::MD5::DES: index
0
1
2
3

As I've mentioned this is only affecting 1 host out of 23
Any help will be appreciated.

Thanks

[tbsh]

Hi,

I can't see why net-snmp is the problem here as I'm getting the CPU cores listed on the command line, but not from within the cacti interface.

Also nothing changes on these servers including the one cacti is running on.

[Dev Null]

I have what seems to be a similar issue with the ucd/net - Get Monitored Partitions query. When I run the query in verbose mode I get:

Code: Select all

+ Found data query XML file at '/var/www/html/cacti/resource/snmp_queries/net-snmp_disk.xml'
+ XML file parsed ok.
+ Executing SNMP walk for list of indexes @ '.1.3.6.1.4.1.2021.9.1.1'
+ No SNMP data returned

But if I query the same OID from the command line on the same machine I get:

Code: Select all

# snmpwalk -v 3 -l AuthNoPriv -u user -A password machine.nau.edu '.1.3.6.1.4.1.2021.9.1.1'
UCD-SNMP-MIB::dskIndex.1 = INTEGER: 1
UCD-SNMP-MIB::dskIndex.2 = INTEGER: 2

Both machines are running NET-SNMP version: 5.3.2.2. There's nothing much interesting in the logs, but /var/log/messages on the monitored host shows:

Code: Select all

Jun  2 13:36:20 mork snmpd[25035]: Received SNMP packet(s) from UDP: [host.ip]:51982 
Jun  2 13:41:19 mork snmpd[25035]: Connection from UDP: [host.ip]:38556 
Jun  2 13:41:19 mork snmpd[25035]: Received SNMP packet(s) from UDP: [host.ip]:38556 

...when I run the query from the command line, while nothing of the sort shows up in the log when I attempt the verbose query from within cacti.