Any one know the OID on the HP9308 switch for monitoring CPU Utilization?

I hope you appreciate this. It's 1.3.6.1.4.1.11.2.14.11.5.1.9.6.1

I had to download a 20 meg demo program called oidview (from
http://www.oidview.com/oidview.html)

and install it then get the MIB pack from HP.

http://www.hp.com/cposupport/swindexes/mibsforhps11110_swen.html

and put it in the MIB folder for oidview and then tell it I wanted the statistics package hpSwitchStat.mib. Of course it then wanted another package but once I showed it where the first one was it went and found all of the others itself.

-- HP Enterprise Switch Statistics MIB


STATISTICS-MIB DEFINITIONS ::= BEGIN

IMPORTS
Counter, IpAddress
FROM RFC1155-SMI
OBJECT-TYPE
FROM RFC-1212
DisplayString
FROM RFC1213-MIB
HpSwitchPortType
FROM HP-ICF-TC
hpSwitch
FROM HP-ICF-OID;

....

hpSwitchCpuStat OBJECT-TYPE
SYNTAX INTEGER (0..100)
ACCESS read-only
STATUS mandatory
DESCRIPTION "The CPU utilization in percent(%)."
::= { hpSwitchMiscStat 1 }

It would have been a lot easier if I'd had one of these switches to play with. Then just load the HPmibs into getif and point it at the switch and walk down to it.

Ron

.iso.org.dod.internet.private.enterprises.hp.nm.icf.hpicfObjects.hpicfSwitch.hpSwitch.hpSwitchStatistics.hpSwitchMiscStat.hpSwitchCpuStat

or

.1.3.6.1.4.1.11.2.14.11.5.1.9.6.1

HP MIBs can be found at http://www.hp.com/rnd/software

To find the MIB / OID in the future you may need to use a SNMPwalk type of program.

Regards,
Sergejs

Thanks guys really appreciate
this...However, now I'm having problem loading the oid into my mrtg graph. Its not taking the oid that you guys provided.

What does your MRTG config file entry look like for the CPU Utilization? Exactly what error are you getting. Did you specify two OIDs on the target line? Did you set the gauge option? Did you set the MaxBytes to 100? Following is from one of my working mrtg.cfg files which measure input and output air temp on a cisco 7500 router:

Target[NELTemp]: 1.3.6.1.4.1.9.9.13.1.3.1.3.1&1.3.6.1.4.1.9.9.13.1.3.1.3.2:public@10.10.10.1

Options[NELTemp]: nopercent,gauge
YLegend[NELTemp]: Degrees C
ShortLegend[NELTemp]: C
Legend1[NELTemp]: Incoming Air Temp
Legend2[NELTemp]: Outgoing Air Temp
MaxBytes1[NELTemp]: 100
MaxBytes2[NELTemp]: 100
Title[NELTemp]: NEL Air Temp
PageTop[NELTemp]:

NEL Air Temp

What does your MRTG run on? If Windows use getif to query the box and see what it gives you for that value.

http://www.wtcs.org/snmp4tpc/getif.htm

If *nix one of the snmpwalk or snmpget programs should do the same.

Run

The error message is "Skipping update inlast and outlast is not defined"

Here's the entire script:

Target[hp9308_cpu]: 1.3.6.1.4.1.11.2.14.11.5.1.9.6.1&1.3.6.1.4.1.11.2.14.11.5.1.9.6.1:public@10.100.1.2
MaxBytes[hp9308_cpu]: 100
Options[hp9308_cpu]: nopercent, gauge, unknaszero
Title[hp9308_cpu]: dw9308-1
PageTop[hp9308_cpu]:

CPU Utilization dw9308-1

YLegend[hp9308_cpu]: CPU Utilization
ShortLegend[hp9308_cpu]: %
LegendI[hp9308_cpu]: CPU
LegendO[hp9308_cpu]: .
Legend1[hp9308_cpu]: CPU Utilization
Legend2[hp9308_cpu]: .
Legend3[hp9308_cpu]: Max value per interval on graph
Legend4[hp9308_cpu]: .
Colours[hp9308_cpu]: GREEN#00eb0c,BLUE#0000ff,GRAY#AAAAAA,VIOLET#ff00ff
WithPeak[hp9308_cpu]: ymw
Unscaled[hp9308_cpu]: dwmy

Try adding a .0 to the OID if that doesn't work try a .1. Could be that the MIB is allowing for the possibility of multiple CPUs.

Ron

I already tried entering both 0 and 1 and still get error message "did not eval into defined data" and " inlast and outlast is not defined."
Any ideas what could be wrong???

Try next OIDs:
1.3.6.1.4.1.11.2.14.11.5.1.9.6.0
or
1.3.6.1.4.1.11.2.14.11.5.1.9.6.1
or
1.3.6.1.4.1.11.2.14.11.5.1.9.6.2

Regards,
Sergejs

0, 1 or 2 didn't work......I don't think these are the right OID's. When I used OIDView Professional, the variable name hpSwitchCPUStat has no data instances which is why it isn't generating any output on my graph.

Do you have the latest software version?

07.6.01b is the latest on the website at:

http://www.hp.com/rnd/software/switches.htm

It just came out last month.

Ron

I'm running 7.504 which shouldn't make a difference. I don't see any fixes for snmp related bugs on 7.6.01b....

So who trusts the release notes. ;-)

Just get the program getif and point it at the beast.

http://www.wtcs.org/snmp4tpc/getif.htm

Load the HP mibs in the Mibs folder under getif. Put in the hostname (best to use the ip address) and community strings and hit start. It should come back with the SysName at least. Then go to MBrowser and open up the iso until you get to the one you want or paste the number in at the top. Then press Walk. If it still doesn't work then either upgrade the software or talk to HP.

Ron

Here are the oids I was able to use for cpu utilization on the HP9308 "1.3.6.1.4.1.11.2.3.7.11.12.1.2.1.35.0" or .iso.org.dod.internet.private.enterprises.hp.nm.system.netElement.hpEtherSwitch.hpSwitch.switch.snAgentSys.snAgentGbl.snAgGblCpuUtilData.
The data type is a gauge and so far the graph is comparable to command line results.

I am working on a school project and trying to monitor CPU usage on a HP procurve 2512 switch, using MRTG. Below is pasted from the config file:
Target[Jalla_cpu]: 1.3.6.1.4.1.11.2.14.11.5.1.9.6.1.0&1.3.6.1.4.1.11.2.14.11.5.1.9.6.1.0:public@10.30.0.10
MaxBytes[Jalla_cpu]: 100
Options[Jalla_cpu]: nopercent, gauge, unknaszero
Title[Jalla_cpu]: HP switch
PageTop[Jalla_cpu]:

HP switch

YLegend[Jalla_cpu]: CPU Utilization
ShortLegend[Jalla_cpu]: %
LegendI[Jalla_cpu]: CPU
LegendO[Jalla_cpu]: .
Legend1[Jalla_cpu]: CPU Utilization
Legend2[Jalla_cpu]: .
Legend3[Jalla_cpu]: Max value per interval on graph
Legend4[Jalla_cpu]: .
Colours[Jalla_cpu]: GREEN#00eb0c,BLUE#0000ff,GRAY#AAAAAA,VIOLET#ff00ff
WithPeak[Jalla_cpu]: ymw
Unscaled[Jalla_cpu]: dwmy

My problem is getting MRTG to make a website with graphs from this. I have no problems getting graphs when I monitor traffic on the ports.
I have tryed the OIDs and gotten response from them.
Since I am rather clueless, I hope someone know what I did wrong. Or didnt do.

It's a bad idea to post a new problem on the end of an old one. Then you can't reward the answers with points so you won't get as many responses.

However, you need to be a little more specific about your problem. What exactly is the problem? Does it create the log files and not do the graphs or does it not create any log files at all or are the log files and graphs all zero? Does it finish without errors? (Run it manually and see if it has any complaints.)

When you tried the OID what value did you get back?

Are you running on UNIX, LINUX or Windows?

Ron

It does not create any log files at all, or other files. That is the big problem sort of.
I get no errors when finishing it.
Values from the OID varied from 6 to 7.
Running Linux, redhat 8.

I run MRTG on an NT box so some things are a bit different. In NT I have a bat file which the scheduler calls. It is nothing more than: perl mrtg mrtg.cfg (tho with full paths for mrtg and mrtg.cfg). If I just run that command I get error messages in the current window. Have you tried that? I've never been able to redirect the messages to another file so I have to sit there and watch it.

Is your cfg file appended to the bottom of the successful cfg file for the traffic or a stand alone? If stand alone you need to set the directory at the top of the file.

WorkDir: /home/http/mrtg

Make sure the directory exists.

Also look for a lock file with an old date on it. On mine it is mrtg.cfg_l. Should be in the same directory as mrtg and mrtg.cfg (the config file). If you find one you need to remove it.

I trust you have the correct community string and ip address for the switch and that the mrtg box is on the same VLAN as the management IP for the switch.

Ron