[SOLVED] Possible query_host_cpu.php issue in 0.8.7a

[skinty]

I have been digging through the posts, but I have been able to resolve one lingering issues. I cannot add CPU graphs or aggregate CPU graphs after my upgrade to 0.8.7 and 0.8.7a.

I have verified my files against the ones posted in the sticky between the transition to 0.8.7a which basically have the updated .php and .xml files above. Here is the output listed from a verbose query:

Code: Select all

+ Running data query [9].
+ Found type = '4 '[script query].
+ Found data query XML file at '/usr/share/cacti/resource/script_queries/host_cpu.xml'
+ XML file parsed ok.
+ Executing script for list of indexes '/usr/bin/php -q /usr/share/cacti/scripts/query_host_cpu.php 192.168.101.10 **** 2 index'
+ Executing script query '/usr/bin/php -q /usr/share/cacti/scripts/query_host_cpu.php 192.168.101.10 **** 2 query index'
+ Found item [hrProcessorFrwID='0'] index: 0
+ Found item [hrProcessorFrwID='1'] index: 1
+ Found item [hrProcessorFrwID='2'] index: 2
+ Found item [hrProcessorFrwID='3'] index: 3
+ Found data query XML file at '/usr/share/cacti/resource/script_queries/host_cpu.xml'
+ Found data query XML file at '/usr/share/cacti/resource/script_queries/host_cpu.xml'
+ Found data query XML file at '/usr/share/cacti/resource/script_queries/host_cpu.xml'

I can re-index just fine, but when I create the graphs, I get the following output:

Code: Select all

+ Created graph: avapps2 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: avapps2 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: avapps2 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: avapps2 - CPU Utilization - CPU|query_hrProcessorFrwID|

This breaks the data source association and graph creation.

TIA for any help.

-ryan

[Smux]

I have a same problem descript in my issue. but my server crashs when i click in verbose query to show information for my processors

[skinty]

Actually,

It looks like your issue might be slightly different, but using a Debian install and patching the install, I end up with your issue on a different box.

My production box returns the following from command line:

Code: Select all

/usr/bin/php -q /usr/share/cacti/scripts/ss_host_cpu.php 192.168.101.10 2 2:161:500:*****:::MD5:DES: index
0
1
2
3

And:

Code: Select all

/usr/bin/php -q /usr/share/cacti/scripts/query_host_cpu.php 192.168.101.10 ***** 2 index
0
1
2
3

It's the next step of creating the data sources and graphs that it bombs on.

[skinty]

Any suggestions on where I should look.

[skinty]

Hi,

I'm still not getting anywhere with this issue. The install is on a Debian etch machine. However, cacti is manually installed.

As mentioned in earlier posts, I can index a single, dual, quad, or 8 processor machine. However, when I try to create the graphs, it doesn't seem to populate the correct information:

Code: Select all

+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|
+ Created graph: *****-ex1 - CPU Utilization - CPU|query_hrProcessorFrwID|

I have been using the multi cpu graphs from the following post http://forums.cacti.net/about11615.html

I tried using a default install of 0.8.7a files to see if it would resolve the issue, but it yields the same results.

Any help would be greatly appreciated.

[Smux]

My other server with RHEL 4 works fine with cacti 0.8.7 and 0.8.7a, but in my server with Centos 5 Final dont work.

[TheWitness]

Net-SNMP Agent needs updating. I have seen this issue and it was resolved with an upgrade to the agent.

TheWitness

[Smux]

TheWitness,

I have updated my net-snmp agent to 5.4.1

Code: Select all

root@svr01 [/home/admin]# /usr/sbin/snmpd -v

NET-SNMP version:  5.4.1
Web:               http://www.net-snmp.org/
Email:             net-snmp-coders@lists.sourceforge.net

But I'm still having the same problem described in this issue.

[skinty]

I tried upgrading from the Debian installed version of 5.3.1 to 5.4.1. Still having the same results.

What type of debug information would be useful to help with troubleshooting?

[skinty]

When I try using the rebuild poller cache, I get the following errors:

Code: Select all

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Notice: Undefined index: script_function in /usr/share/cacti/lib/utility.php on line 168

Not sure if it's related.

[skinty]

I did a fresh install on another box that is also running Debian etch. Everything is hand installed, so I'm not using the maintainer version. I imported the multi CPU graphs again and indexed a server, which now produced the numbered CPU graphs (0-3):

Code: Select all

+ Created graph: ********** - CPU Utilization - CPU0
+ Created graph: ********** - CPU Utilization - CPU1
+ Created graph: ********** - CPU Utilization - CPU2
+ Created graph: ********** - CPU Utilization - CPU3

It seems that during imports of different versions of the CPU script, I managed to bork my data input query for snmp. Exporting from my new system and importing into the broken system fixed the graph creation. However, it wasn't until I rebuilt the poller cache that everything else was fixed.

php /usr/share/cacti/cli/rebuild_poller_cache.php and now all is well, including my Advanced Ping templates (which stemmed back to multiple cron entries with different user names)

[TheWitness]

So, a bad distribution then? Did you open a ticket with the maintainer?

TheWitness

[exploit-soget]

I have the same problem on a SLES10 SP1 O.S with manual installation of Cacti (0.8.7a).

When i try Get Processor Information : i have 0 items and 0 rows.

Can anyone help me ?

[tonzy0907]

My other server with RHEL 4 works fine with cacti 0.8.7 and 0.8.7a, but in my server with Centos 5 Final dont work.

Smux,

We have the same problem. Did you get a solution on this?

[gandalf]

I was tired of this.
And I in fact do not like using scripts if pure SNMP is all that's needed. That's why I decided to publish s PURE SNMP replacement of the CPU thingy. It's fast. It uses some nice features of cacti. And it works (at least for me).
See Scripts and Templates forum.
Reinhard