[solved] Empty graphs only for Windows Filesystems

[cnomark]

Hello,
I've got a problem with windows Devices.
Cacti does not for an WinXP SP3 Host display the used Memory (Ram and Swap) and not display the Used(unused space from Harddisks.

In the log I see this for the the host:

05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[642] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get used 6, output: 373227520
05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[642] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get total 6, output: 2138505216
05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[641] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get used 5, output: 251920384
05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[641] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get total 5, output: 3055091712
05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[640] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get total 4, output: 1999382315010
05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[640] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get used 4, output: 1901379141630
05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[639] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get used 2, output: 39784390656
05/05/2010 03:00:38 PM - CMDPHP: Poller[0] Host[29] DS[639] SERVER: /usr/share/cacti/site/scripts/ss_host_disk.php ss_host_disk cnopcs04 29 1:161:600:1:10:public:::::: get total 2, output: 0

------
But in the graph appears only 0.0 for everything.
What happens here? Everything else works fine.

Thx, for your help!

Technical information

General Information
Date Wed, 05 May 2010 14:41:19 +0200
Cacti Version 0.8.7e
Cacti OS unix
SNMP Version NET-SNMP version: 5.4.1
RRDTool Version RRDTool 1.3.x
Hosts 12
Graphs 205
Data Sources Script/Command: 4
SNMP: 93
SNMP Query: 79
Script Query: 66
Script Query - Script Server: 41
Total: 283
Poller Information
Interval 300
Type cmd.php
Items Action[0]: 248
Action[1]: 164
Action[2]: 68
Total: 480
Concurrent Processes 1
Max Threads 1
PHP Servers 1
Script Timeout 25
Max OID 10
Last Run Statistics Time:84.6113 Method:cmd.php Processes:1 Threads:N/A Hosts:12 HostsPerProcess:12 DataSources:375 RRDsProcessed:205
PHP Information
PHP Version 5.2.6-1+lenny8
PHP OS Linux
PHP uname Linux cnong01 2.6.26-2-686 #1 SMP Tue Mar 9 17:35:51 UTC 2010 i686
PHP SNMP Not Installed
max_execution_time 30
memory_limit 128M

[BSOD2600]

I don't think XP supports either of those OIDs... anyways snmpwalk .1.3.6.1.2.1.25.2.3.1 to manually check. If it does, then try increasing the snmp timeout to that device by 2-3 seconds.

[cnomark]

Hhmmmm....
I had here today in the morning an reply posted in that I have wrote, that the solution from BSOD2600 not helped.
Now it's away!?

OK, I have found an solution:

I have the Host-Template Windows 2000/XP Host compared with an new and clean install of Cacti, because the cacti with the problem was an update from an older one.

I have seen, that in the new one the Host-Tamplate in the "Associated Data Queries" section for the Filesystem the dataquery "SNMP - Get Mounted Partitions" has.

In the old one was there ucd 10 fs.
So I have the Data query from the new one exported and in the old Cacti imported. After that I have the graphs regenerated.

Now it works!


Thanks
cnomark

[stalingrad89]

Hello,

I have kind of the same problem.
My Cacti is installed on a Debian.
I try to monitor a Windows server 2008.
But the graphs are empty, they are always to 0.0 or nan whatever the graph is.

I noticed that my associated queries seem to be the same as you cnomark.
What are exactly the associated queries you use?
What is the version of Cacti you have installed ?

Thanks

[cnomark]

....
What are exactly the associated queries you use?
What is the version of Cacti you have installed ?

Thanks

I have installed Version 0.8.7e with all patches.
Te query is a self written script (perl and xml) as Data Input Method Get Script Data (Indexed)

Take a look in the log of cacti.
Set Poller Logging Level to devel.
Mostly you can see there the reason.

Try also the Debug-Mode of the Graphs.


cnomark

[GreenMan24]

Hi All,

i have the same problem with certain windows 2008 servers, where the HDD appeared normally on some server hosts where not in other (2 failure hosts from 29 total) also the CPU appeared normally on all the servers except one, what i want to say that the cacti can read some data normally from server and not from the other, although both of them work on the same OS.

i tried to search for this problem, and found that this topic is the most close to mine

also i used most of the tamplates for the windows .

please help.

below the technical info:
cacati installed version: Version 0.8.7e
installed on server windows 2003 ( Dell Box_optilix 330)
monitured servers added to the cacti (solaris, windows server 2008, and CentOS)

also attached,
the empty graph debug
normally graph debug
cacti log file

Thanks,
GreenMan24

[BSOD2600]

please follow http://forums.cacti.net/viewtopic.php?t=11747 for additional troubleshooting. All of those "Result from SNMP not valid. " need to get fixed.

[GreenMan24]

Thanks,

first for all the mentioned installation procedure is the one i used, also the technical support is attached, but regarding the graphs just as i mentioned earlier all the graphs appeared normally on the cacti except little of them ( actually for the servers only, and here i want to ask that there is any action must be done on the servers to get the data, in other words the problem may from the server itself because other servers with the same type have no problems)
and regarding the warning appeared on the cacti log file ( WARNING: Result from SNMP not valid. Partial Result: U) i searched for this warning and found that it should be solved by changing the Re-Index Method from _Up time Goes backward_ to _verify all fields_ but this action actually do not clear this warning but add new one which is (WARNING: Re cache Event Detected for Host)

regarding the debugging of the graphs, all the mentioned procedure can be done on the UNIX machine but unfortunately the cacti is installed on windows machine, i don't know if it can be checked on the windows machine.

please let me know if you can open the web page normally.

Thanks Again
GreenMan24

[BSOD2600]

...and regarding the warning appeared on the cacti log file ( WARNING: Result from SNMP not valid. Partial Result: U) i searched for this warning and found that it should be solved by changing the Re-Index Method from _Up time Goes backward_ to _verify all fields_ but this action actually do not clear this warning but add new one which is (WARNING: Re cache Event Detected for Host)

That's not likely the correct fix, as you never ascertained what the real problems were. You shouldve changed the logging level to high/debug and then found out what queries are failing and went from there. The recache warning is fine/expected.

regarding the debugging of the graphs, all the mentioned procedure can be done on the UNIX machine but unfortunately the cacti is installed on windows machine, i don't know if it can be checked on the windows machine.

Not all commands are for unix... and even so, any IT person will know the equivalent Windows command to accomplish the same task.