AIX Host Template

hootjr29 · Post by **hootjr29** » Wed Feb 23, 2005 1:30 pm

sorry... forgot to mention that I unzipped the archive and imported the template without any errors. Everything said successful. Then, I went to add one of my aix hosts using this template. THat is when I got those errors above.

ALso, as far as the interfaces go, I didn't realize this, but the Generic SNMP template has the interfaces that seem to work fine for graphing network traffic. So, I think once I can get your template working good, I will be all set, since all I really care about is traffic and filesystem space... maybe in the future I'll do CPU and Users.

Post by **rony** » Wed Feb 23, 2005 1:57 pm

Did you copy the xml files from the resource directory in the zip file to the respective resource directory in your cacti installation? That should fix that problem.

As for CPU usage, you can add the Data Query "SNMP - Get Processor Information" to the host manually or if you are doing a lot, you can add it to the Host Template. This will give you the Host mib graphs, which will show you each CPU. I will look at the AIX CPU mib and see, I think it's a summation of all the CPU's but I'm not completely sure.

Post by **rony** » Wed Feb 23, 2005 2:30 pm

Checked into it, I can't find any CPU references in the AIX mib.

Also, has anyone has any success graphing hrSystemProcesses and hrSystemNumUsers from the HOST mib? They are standard graph templates in Cacti. For AIX 5.2, I consistantly get a warning that the type is wrong which doesn't make cacti, and most snmp query tools happy.

Anyone not having this problem, can you please post your oslevel -r.

Thanks,

hootjr29 · Post by **hootjr29** » Thu Feb 24, 2005 8:22 am

mib/oid that gives you total users logged in:
This is in /usr/samples/snmpd/hr.my

Code: Select all

snmpinfo -m dump -v | grep -i hrSystemNumUsers.0
snmpwalk -c public -v 1 aixservernamehere 1.3.6.1.2.1.25.1.5.0

mib/oid that gives you a summation percentage of all cpu's:
This is in /usr/samples/snmpd/aixmib.my

Code: Select all

snmpinfo -m dump -v | grep -i aixSeCPUUtilization
snmpwalk -c public -v 1 aixservernamehere 1.3.6.1.4.1.2.6.191.1.2.1.0

Also, Rony, for some reason, I can use every template as expected with the following exceptions:

1) Can't use Volume Group Usage at all.
When I look through the cacti interface at Graph Templates, I have the following:

AIX File System Statistics
AIX File System Usage
AIX Paging Volume Usage
AIX Physical Volume Usage
AIX Volume Group Statistics

But no AIX Volume Group Usage. I see it in the template.xml that you've created. However, I noticed that there is no hash information for it in the graph_template_id. There is currently a zero instead of a hash. Do you know what I might need to do in order to get that graph template?

2) The page space displays just fine. However, when I click on it, the additional trending graphs (daily, weekly, monthly, yearly) don't display. Any idea what I might need to do there?

3) The same thing that happens with Page Space happens with File System Statistics. I won't be using this one, so I'm not concerned with it, but I thought I'd at least bring it up for troubleshooting purposes.

Thanks again for all of your hard work,
Joe

Post by **rony** » Thu Feb 24, 2005 8:38 am

hootjr29 wrote:mib/oid that gives you total users logged in:
snmpinfo -m dump -v | grep -i hrSystemNumUsers.0
snmpwalk -c public -v 1 aixservernamehere 1.3.6.1.2.1.25.1.5.0

I know the OID, my problem is that the SNMP daemon is returning an Interger not a gauge for that OID. I have a support PMR open with IBM concerning it.

Concerning the CPU usage, I will add that template, I totally missed that.

I will look into why the template didn't include that hash correctly. And re-export the templates.

hootjr29 · Post by **hootjr29** » Thu Feb 24, 2005 9:07 am

yeah... I noticed that about the users as well. Hopefully IBM can help us.

Oh... and for anyone who is using AIX 5.3. Have you had any luck with using SNMP? If I replicate the procedure that I mentioned earlier as far as getting SNMP up and running so that you can query it, I don't get the same results on AIX 5.3 servers.

Post by **rony** » Thu Feb 24, 2005 9:25 am

What results are different between 5.3 and 5.2?

Have you contacted IBM about it?

They are getting tired of hearing from me.... I keep talking to the 2 same people in the same area about snmp.

Attached you will find an updated template that has corrected the Volume Group Usage link and now includes the AIX CPU Utilitzation.

Edit: Attachment removed, later fixed version in post.

hootjr29 · Post by **hootjr29** » Thu Feb 24, 2005 10:10 am

I'll try it out. I have contacted IBM a bunch too. Let me test a little more to verify that I haven't done something wrong on the AIX 5.3 side of things. But when I went through an initial test of just modifying the /etc/snmpdv3.conf file and including the aixmibd portion, restarted snmpd and all related services, I can't query some of the same things that I could query from AIX 5.2, such as aixFs or hrStorage. So its almost like public can't see those mibs.

I'll try again and keep you informed.

Post by **rony** » Thu Feb 24, 2005 11:00 am

Pay close attention to the exclude and include lines in the configuration file. They are proccessed in the order they appear in the file. So, if you have an exclude before you have an include, the exclude is applied.

Also, the order which you stop and start the daemons is important. Remember to start snmpd then the sub agents. When stopping them, don't forget to make sure they exit fully before restarting.

Personally go bit by both of these.

hootjr29 · Post by **hootjr29** » Fri Feb 25, 2005 9:55 pm

I'll try again within the next couple of weeks. I ran into an issue where SNMP is causing some issues with oracle. It seems that oracle has a listener or something that is logging tons of errors in the event log because of SNMP. I've disabled it on 1 of 2 AIX servers that I'm currently working with. But the template is working wonderfully with AIX 5.2. I will continue testing AIX 5.3 within the next couple of weeks...

Thanks again Tony!

joe

pbulteel · Post by **pbulteel** » Mon Feb 28, 2005 1:19 pm

Cool! Except that the import works, but I don't get anything...

I get a "NO SNMP data returned." -- see output below.

Code: Select all

+ Running data query [10].
+ Found type = '3' [snmp query].
+ Found data query XML file at '/var/www/html/cacti/resource/snmp_queries/aix_vg.xml'
+ XML file parsed ok.
+ Executing SNMP walk for list of indexes @ '.1.3.6.1.4.1.2.6.191.2.1.2.1.9'
+ No SNMP data returned
+ Found data query XML file at '/var/www/html/cacti/resource/snmp_queries/aix_vg.xml'
+ Found data query XML file at '/var/www/html/cacti/resource/snmp_queries/aix_vg.xml'
+ Found data query XML file at '/var/www/html/cacti/resource/snmp_queries/aix_vg.xml'

I'm going to clean the system and remove the templates... make sure I don't have any of the old templates, code or xml and then try to reimport this.

Post by **rony** » Tue Mar 01, 2005 7:55 am

Refer to the first post I made.

Make sure that you are running the aixmibd deamon and that you have configured snmpd to include the .1.3.6.1.4.1.2.6.191 OID in it's tree.

pbulteel · Post by **pbulteel** » Wed Mar 02, 2005 1:30 pm

Duh!!! I thought I had done that, but I just noticed that the line that was in there said exclude. So I changed it to include and voila!

Thank you!

-P

Post by **rony** » Wed Mar 02, 2005 2:36 pm

Yah!

That one took me a while to find when I was working on it.

pbulteel · Post by **pbulteel** » Thu Mar 03, 2005 9:48 am

I think I found a bug in cacti - probably the import.

What happened was my INodes graph (File System Statistics) was graphic correctly in preview mode but now when you went to view it I would get an error.

When I debugged I found that the problem was with the GPRINT.

Code: Select all

 /usr/bin/rrdtool graph - --imgformat=PNG --start=-86400 --end=-300 --title="Ealcsaix03 - Argos test environment - File System Statistics - /dev/hd10opt" --base=1024 --height=120 --width=500 --alt-autoscale-max --lower-limit=0 --vertical-label="INodes" DEF:a="/var/www/html/cacti/rra/ealcsaix03_argos_test_environment_aixfsnuminodes_295.rrd":aixFsNumINodes:AVERAGE DEF:b="/var/www/html/cacti/rra/ealcsaix03_argos_test_environment_aixfsnuminodes_295.rrd":aixFsUsedInodes:AVERAGE AREA:a#00004D:"Total"  GPRINT:a:LAST:"Current\:"  GPRINT:a:MIN:"Minimum\:"  GPRINT:a:MAX:"Maximum\:\n"  AREA:b#CAF100:"Used"  GPRINT:b:LAST:"Current\:"  GPRINT:b:MIN:"Minimum\:"  GPRINT:b:MAX:"Maximum\:\n"  COMMENT:"Mount Point: /opt\n"

It turns out that the Current, Max, etc are missing the %8.0lf and are therefore failing.

FIX IN CACTI - NOT CODE

I went to the console, opened up the graph template. I edit the graph item and save. I don't make any changes, just open and save. When this is complete you can then view all the graphs that use this template.

-P

AIX Host Template

Who is online