Query multiple SNMP OID's no index

[Michalis]

I have been reading any documentation i could find and searched the forums extensively but i am not sure if this is possible or not.
Basically i want to use an SNMP query with multiple OID's with no index.
An example of this is getting the memory information from SNMP informant.
Can i do this with the following query?
Note : I put the oid_index stuff to make the parsing work with cacti.

Code: Select all

<query> 
<name>Win32 - Memory Statistics</name> 
<description>Grab the memory statistics using SNMP-Informant</description> 
<oid_index>.1.3.6.1.4.1.9600.1.1.2</oid_index>
<index_order_type>oid</index_order_type> 

<fields> 
	<AvailableBytes> 
	<name>Available Bytes</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.1</oid>
	</AvailableBytes> 
	
	<AvailableMBytes> 
	<name>Available Megabytes</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.3</oid>
	</AvailableMBytes> 

</fields> 
</query>

If this is not the correct way to do it, whats the simplest way to do it without creating external scripts or anything? I like the structured way that is used for indexed queries and i want to use something similar.

Thanks!

[oxo-oxo]

Try it and post verbose query

[Michalis]

Hi oxo-oxo,

I have tried it.
The verbose query shows 0 items, 0 rows

Output is :

Code: Select all

+ Running data query [30].
+ Found type = '3' [snmp query].
+ Found data query XML file at 'C:/apache/Apache2/htdocs/cacti/resource/snmp_queries/snmp_informant_standard_memory.xml'
+ XML file parsed ok.
+ Executing SNMP walk for list of indexes @ '.1.3.6.1.4.1.9600.1.1.2'
+ Found data query XML file at 'C:/apache/Apache2/htdocs/cacti/resource/snmp_queries/snmp_informant_standard_memory.xml'
+ Found data query XML file at 'C:/apache/Apache2/htdocs/cacti/resource/snmp_queries/snmp_informant_standard_memory.xml'
+ Found data query XML file at 'C:/apache/Apache2/htdocs/cacti/resource/snmp_queries/snmp_informant_standard_memory.xml'

No datasources are created.
Is this the correct approach to create a multiple OID query with no index?

[oxo-oxo]

Post snmpwalk from command line of .1.3.6.1.4.1.9600.1.1.2

[Michalis]

Code: Select all

SNMPv2-SMI::enterprises.9600.1.1.2.1.0 = Gauge32: 205901824
SNMPv2-SMI::enterprises.9600.1.1.2.2.0 = Gauge32: 201076
SNMPv2-SMI::enterprises.9600.1.1.2.3.0 = Gauge32: 196
SNMPv2-SMI::enterprises.9600.1.1.2.4.0 = Gauge32: 1593229312
SNMPv2-SMI::enterprises.9600.1.1.2.5.0 = Gauge32: 453611520
SNMPv2-SMI::enterprises.9600.1.1.2.6.0 = Gauge32: 532684800
SNMPv2-SMI::enterprises.9600.1.1.2.7.0 = Gauge32: 9
SNMPv2-SMI::enterprises.9600.1.1.2.8.0 = Gauge32: 0
SNMPv2-SMI::enterprises.9600.1.1.2.9.0 = Gauge32: 0
SNMPv2-SMI::enterprises.9600.1.1.2.10.0 = Gauge32: 0
SNMPv2-SMI::enterprises.9600.1.1.2.11.0 = Gauge32: 33914880
SNMPv2-SMI::enterprises.9600.1.1.2.12.0 = Gauge32: 26783744
SNMPv2-SMI::enterprises.9600.1.1.2.13.0 = Gauge32: 16060416
SNMPv2-SMI::enterprises.9600.1.1.2.14.0 = Gauge32: 435466240
SNMPv2-SMI::enterprises.9600.1.1.2.15.0 = Gauge32: 8192
SNMPv2-SMI::enterprises.9600.1.1.2.16.0 = Gauge32: 1343488
SNMPv2-SMI::enterprises.9600.1.1.2.17.0 = Gauge32: 2076672
SNMPv2-SMI::enterprises.9600.1.1.2.18.0 = Gauge32: 2236416

[oxo-oxo]

So maybe something like this ...:

Code: Select all

<oid_index>.1.3.6.1.4.1.9600.1.1.2.1</oid_index>

The trailing 0 gets turned to Index 1 in the code ...

[TheWitness]

Data Queries "Require" an index. You should instead create a multiple DS Data Template the have each tab in the Data Template have it's own OID.

That way 1 RRD file, multiple OID's and still using snmp.

TheWitness

[BSOD2600]

You should instead create a multiple DS Data Template the have each tab in the Data Template have it's own OID.

Unless something has changed in the past several versions, I thought a DS data template could only do one OID. At least in the past when I've tried to use multiple OID's for a DS template, it only saved one...

[TheWitness]

Well, I thought the same until I saw someone doing it

TheWitness

[Michalis]

I have tried the multiple DS approach per data template in the past but on each DS tab the same OID was showing which lead me to believe that I cannot put multiple DS on a data template. Is there a trick to making this work?

Is there a specific version that this feature was implemented?
I am using cacti-0.8.6j

[TheWitness]

Well don't I feel stupid. Well it should

TheWitness

[Michalis]

OK i think i have figured it out, i am currently graphing the memory information of a server running SNMP informant.
I dont think the memory info has an SNMP index but the following query works and gives me all the memory related information in one RRD which then i can create graphs from.
This is the SNMP query i am using. Note that this is an indexed query.

Code: Select all

<query> 
<name>Win32 - Memory Statistics</name> 
<description>Grab the memory statistics using SNMP-Informant</description> 
<oid_index>.1.3.6.1.4.1.9600.1.1.2</oid_index>
<index_order_type>oid</index_order_type> 

<fields> 
	<memory> 
	<name>Counter Type</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>input</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2</oid>
	</memory>
	
	<AvailableBytes> 
	<name>Available Bytes</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.1</oid>
	</AvailableBytes> 
	
	<AvailableMBytes> 
	<name>Available Megabytes</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.3</oid>
	</AvailableMBytes> 

	<CommitedBytes> 
	<name>Commited Bytes</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.4</oid>
	</CommitedBytes> 

	<CacheBytes> 
	<name>Cache Bytes</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.5</oid>
	</CacheBytes> 
	
	<PageFaultspersec> 
	<name>Page faults per sec</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.7</oid>
	</PageFaultspersec> 	
	
	<PagesInputPerSec> 
	<name>Page inputs per sec</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.8</oid>
	</PagesInputPerSec>	
	
	<PagesOutputPerSec> 
	<name>Page outputs per sec</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.9</oid>
	</PagesOutputPerSec>	
	
	<PagesPerSec> 
	<name>Pages per sec</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.10</oid>
	</PagesPerSec>	

	<PoolPagedBytes> 
	<name>Pool paged bytes</name> 
	<method>walk</method> 
	<source>value</source> 
	<direction>output</direction> 
	<oid>.1.3.6.1.4.1.9600.1.1.2.12</oid>
	</PoolPagedBytes>	

</fields> 
</query>

I guess that this approach shouldn't have worked because there is no SNMP index as such but by using the memory OID as input i can get the information that i need.

[TheWitness]

Right, you get a data query with one line That works for now...

TheWitness

[Michalis]

Well, what about the much fabled mutli-DS data template?
Is this beast at all possible?
I've seen a post in the forum that it shouldn't be allowed for some reason or another.

Still cant figure out a way to do it, all DS tabs share the same OID

Cheers!

[TheWitness]

Yea, I'm going to look at it. But the next release of Cacti is feature frozen at the moment, so it won't be out for a while, and it would be an awfull coplicated and risky patch. So, the DQ method may be the best way to go for now.

TheWitness