Fighting with hard drive polling

[RabidWombat]

I went through the process of moving a cacti installation on Windows to FreeBSD. All devices were polling correctly on the Windows installation. I moved over all of the rrd files and mysql database. Windows was using poller.php, and I left the settings in FreeBSD the same to keep things simple.

All of the graphs except for the harddrives worked when I began polling with the new installation. The computers being polled are Windows computers.

I began digging through the log and discovered that cacti is updating every hard drive rrd with hdd_total:0 and hdd_used:0. I ran scripts/ss_host_disk.php with the same parameters and I received a valid result for all of the hosts I tested. Ok... I ran cmd.php manually, and I received some results, but still mostly zeros. I bumped up the SNMP timeout to 20000 ms and refreshed the poller cache. I saw a few blips of data show up in the hard drive graphs, but nothing consistent. I switched the poller to cactid, and now I start seeing some results. The hard drives are now graphing except for the first one on every device.

I have seen several posts on Windows computers timing out when polling harddrives. However, this doesn't seem to fit. My SNMP timeout is 20 seconds and cactid is finishing all polling around ~15 seconds.

Is there something I'm missing?

[RabidWombat]

After some debugging, I found out that the call to cacti_snmp_get in ss_host_disk.php did not match the arguments of the function.... and that I was 10 days late discovering this

script_server_snmp_auth.patch fixes the problem.

[rony]

Good to hear..

[viss]

RabidWombat - could you post the before/after of your ss_host_disk.php?

I'm having the exact same problem, but on 0.8.6h. Everything works but the disk graphs.

Or, can you point me to where I can find that patch?

[fmangeant]

Or, can you point me to where I can find that patch?

Hi

you can find pre-patches files for 0.8.6h here : http://www.cacti.net/downloads/patches/ ... e-patched/

[Yukata]

Hello everyone.

I'm having the same type of problem with the hard drive space graphs but on a stock Kernel 2.6.15.4 with Fedora Core 2. Here are my tools versions :

rrdtool 1.0.50-2.1
net-snmp 5.1.1-2
MySQL 4.0.20
cacti 0.8.6h
cactid 0.8.6g

The patch doesn't fix this problem and I get a lot of error messages in the logs. Here is a sample of the cacti log in debug mode :

04/17/2006 09:35:00 PM - CACTID: Poller[0] DEBUG: MySQL Query ID '25': 'SELECT

action,hostname,snmp_community,snmp_version,snmp_username,snmp_password,rrd_name,rrd_path,arg1,arg2,arg3,local_data

_id,rrd_num,snmp_port,snmp_timeout FROM poller_item WHERE host_id=2 ORDER BY arg1'
04/17/2006 09:35:00 PM - CACTID: Poller[0] DEBUG: MySQL Query ID '25': OK
04/17/2006 09:35:00 PM - PHPSVR: Poller[0] DEBUG: INCLUDE:

'/usr/local/apache/sites/CACTI/htdocs/scripts/ss_host_disk.php' SCRIPT: 'ss_host_disk' CMD: '192.168.X.X 2

1:161:500:public:::MD5::[None] get total 4'
04/17/2006 09:35:00 PM - CACTID: Poller[0] Host[2] DS[17] SS[0] WARNING: Result from SERVER not valid. Partial

Result:
Warning: main(/usr/...
04/17/2006 09:35:00 PM - CACTID: Poller[0] Host[2] DS[17] SS[0] SERVER:

/usr/local/apache/sites/CACTI/htdocs/scripts/ss_host_disk.php ss_host_disk 192.168.X.X 2

1:161:500:public:::MD5::[None] get total 4, output: U
04/17/2006 09:35:00 PM - PHPSVR: Poller[0] WARNING: Function does not exist

04/17/2006 09:35:00 PM - CACTID: Poller[0] Host[2] DS[17] SS[0] WARNING: Result from SERVER not valid. Partial

Result:
Warning: main(): Fa...
04/17/2006 09:35:00 PM - CACTID: Poller[0] Host[2] DS[17] SS[0] SERVER:

/usr/local/apache/sites/CACTI/htdocs/scripts/ss_host_disk.php ss_host_disk 192.168.X.X 2

1:161:500:public:::MD5::[None] get used 4, output: U
04/17/2006 09:35:00 PM - PHPSVR: Poller[0] DEBUG: INCLUDE:

'/usr/local/apache/sites/CACTI/htdocs/scripts/ss_host_disk.php' SCRIPT: 'ss_host_disk' CMD: '192.168.X.X 2

1:161:500:public:::MD5::[None] get used 4'
04/17/2006 09:35:00 PM - PHPSVR: Poller[0] WARNING: Function does not exist

Does anyone knows where those errors come from ?