Roth, Sandro | 16 Apr 11:42 2014

Missing filesystems with snmpd

Hi all

We have a few veritas clusters running on RHEL5 and 6.
These clusters have veritas file systems on a shared storage.
On RHEL6 I'm able to query these file systems with snmp and graph them with cacti but not on RHEL5.
All servers have the same snmpd.conf.

RHEL6:

$ snmpwalk -v1 -c public localhost hrStorage | grep Desc
HOST-RESOURCES-MIB::hrStorageDescr.31 = STRING: /
HOST-RESOURCES-MIB::hrStorageDescr.35 = STRING: /dev/shm
HOST-RESOURCES-MIB::hrStorageDescr.36 = STRING: /boot
HOST-RESOURCES-MIB::hrStorageDescr.37 = STRING: /home
HOST-RESOURCES-MIB::hrStorageDescr.38 = STRING: /opt
HOST-RESOURCES-MIB::hrStorageDescr.39 = STRING: /tmp
HOST-RESOURCES-MIB::hrStorageDescr.40 = STRING: /var
HOST-RESOURCES-MIB::hrStorageDescr.42 = STRING: /dev/vx
HOST-RESOURCES-MIB::hrStorageDescr.46 = STRING: /global/nw
HOST-RESOURCES-MIB::hrStorageDescr.47 = STRING: /global/aftd

df:

/dev/vx/dsk/dg/nsrvol
                      /global/nw
/dev/vx/dsk/dg/aftdvol
                      /global/aftd

RHEL5:

(Continue reading)

Eric Cables | 9 Apr 06:25 2014
Picon

Boost bug?

The below assumes Boost 5.1 (the latest that I can find)

I'm hoping someone can confirm whether or not I'm hitting a bug with Cacti
and/or Boost. Basically, when Boost creates an RRD file it does not do so
in the same manner as the Data Source debug output, and the resulting RRD
file does not contain correct maximum values for traffic_in & traffic_out.

I posted this thread on the forum (more context, screenshots, and examples):
http://forums.cacti.net/viewtopic.php?f=21&t=50420&p=242799#p242799

Here's an example of the traffic_in & traffic_out max values after Boost
runs, and an RRD file is populated with the data from the MEMORY database:

$ rrdtool info /usr/local/cacti-0.8.8a/rra/57/3407.rrd | grep -i max
ds[traffic_in].max = NaN
ds[traffic_out].max = NaN

When I disable Boost, and re-create the same graph, the RRD file is created
properly:
$ rrdtool info 5175.rrd | grep max
ds[traffic_in].max = 1.0000000000e+10
ds[traffic_out].max = 1.0000000000e+10

Please let me know if I've overlooked something, or if this is a known
limitation. I've run Boost in the past, on a different installation, but
haven't ever run into this problem. I'll add that these are 10GE interfaces
(ifHighSpeed), and the 5.1 changelog lists "feature: Support for
ifHighSpeed."

Thanks,
(Continue reading)

Flur Blubr | 8 Apr 12:17 2014
Picon

Monitoring via ping only, false positive in Manage plugin, old Cacti, looking for patch

I just reallized that images was filtered out. Here are the links to the
pictures:

https://imageshack.com/i/nd1xlep

https://imageshack.com/i/nswiwop

Any idea where the problem can be?

Regards

Fblubr
------------------------------------------------------------------------------
Put Bad Developers to Shame
Dominate Development with Jenkins Continuous Integration
Continuously Automate Build, Test & Deployment 
Start a new project now. Try Jenkins in the cloud.
http://p.sf.net/sfu/13600_Cloudbees
Flur Blubr | 7 Apr 08:39 2014
Picon

Monitoring via ping only, false positive in Manage plugin, old Cacti, looking for patch

Dear list,

please help me understand what is wrong with our older installation - Cacti
0.8.7g. We can monitor few backup wirewalls only by ping, so I disabled
SNMP options and Cacti is doing only TCP ping to the device. Surely this
works and they are green in Cacti console, howewer our SPOC team is using
the Manage plugin view where all thoose firewalls looks down. So no way to
detect when device is not responding to pings since it looks like down all
the time.

While searching where the problem can be I installed newer version of
Manage plugin to Cacti 0.8.8a and detected the behaviour is normal there,
devices monitored by ping only are green and looks up in Manage plugin
view. Is this some well known bug fixed in never versions so probably I can
find a patch even for my old intallation? We are not allowed to upgrade
this Cacti. Two screenshots attached for better understanding.

Please share your opinion

Regards

Fblubr
------------------------------------------------------------------------------
Put Bad Developers to Shame
Dominate Development with Jenkins Continuous Integration
Continuously Automate Build, Test & Deployment 
Start a new project now. Try Jenkins in the cloud.
http://p.sf.net/sfu/13600_Cloudbees_APR
(Continue reading)

trunk | 4 Apr 21:14 2014
Picon

(no subject)

Hi all! I found bug in leatest stable version cacti 0.8.8b was installed today.

When i making graphs from http://forums.cacti.net/viewtopic.php?f=12&t=11020 all works fine without
one piece.
When i opening graph preview of host with that graph, my preview was not display images, but if i open link of
this broken image then graphs works fine.

After debug function i was found bug place and fix it.

 patch for file lib/rrd.php

 <at>  <at>  -1347,14 +1347,14  <at>  <at> 
                                 $comment_string = $graph_item_types{$graph_item["graph_type_id"]} . ":" . str_replace(":", "\:",
cacti_escapeshellarg($graph_variables["text_format"][$graph_item_id] .
$hardreturn[$graph_item_id])) . " ";
                                 if (trim($comment_string) == 'COMMENT:"\n"') {
                                         $txt_graph_items .= 'COMMENT:" \n"'; # rrdtool will skip a COMMENT that holds a NL only; so add a blank to
make NL work
- } else if (trim($comment_string) != "COMMENT:\"\"") {
+ } else if ((trim($comment_string) != "COMMENT:\"\"")&&(trim($comment_string) != "COMMENT:''")) {
                                         $txt_graph_items .= rrd_substitute_host_query_data($comment_string, $graph, $graph_item);
                                 }
                         }else {
                                 $comment_string = $graph_item_types{$graph_item["graph_type_id"]} . ":" .
cacti_escapeshellarg($graph_variables["text_format"][$graph_item_id] .
$hardreturn[$graph_item_id]) . " ";
                                 if (trim($comment_string) == 'COMMENT:"\n"') {
                                         $txt_graph_items .= 'COMMENT:" \n"'; # rrdtool will skip a COMMENT that holds a NL only; so add a blank to
make NL work
- } else if (trim($comment_string) != "COMMENT:\"\"") {
(Continue reading)

Rendy Guo | 29 Mar 12:02 2014
Picon

question about between OS and cacti

Hello everyone.

I'm a new CACTI user.

I want to know that which OS is suitable for CACTI.
Is CACTI installed in windows will be unstable?

Thanks in advance.

Regards.
------------------------------------------------------------------------------
Paolo De Michele | 13 Mar 20:56 2014
Picon

no graph data received

hi everybody,

I followed this url [1] and I have added just a server to test the
monitoring
Actually after 15 minutes I can't see any graph

Some detail:
CentOS 6.4 x32
Cacti 0.8.8b (epel repository)
rrdtool 1.3.8-6

In device's section the status of machine is alway "unknown"
but if I explore

monitoring.x.local (172.16.x.x)
SNMP Information
*System:*1:56:33:UTC:2014:I686
*Uptime:* 92545 (0 days, 0 hours, 15 minutes)
*Hostname:* monitoring
*Location:* Unknown
*Contact:* root <at> localhost

How can I fix it?
Thanks in advance

Regards

[1]
http://www.tecmint.com/install-cacti-network-monitoring-on-rhel-centos-6-3-5-8-and-fedora-17-12/
------------------------------------------------------------------------------
(Continue reading)

Ian Bennett | 28 Feb 11:46 2014
Picon

XML files

Evening all,
	I've been tinkering with snmp queries, writing XML files but have struck a snag.
	All of what I've done so far has used a single index to gather data from multiple instances, but
what if I need to gather data using different indexes??
	Here is what I'm trying to do. We have a number of Cisco 6500 switches in our network and I want to
collect CPU data for graphing. Depending on the types of modules installed in the slots, will depend
on how many CPUs are available to be interrogated.
	All the CPU info is under the base OID of .1.3.6.1.4.1.9.9.109.1.1.1.1
	The following is a walk of that OID from one of our switches:
.1.3.6.1.4.1.9.9.109.1.1.1.1.2.1 = 2017  #cpmCPUTotalPhysicalIndex
.1.3.6.1.4.1.9.9.109.1.1.1.1.2.2 = 2001
.1.3.6.1.4.1.9.9.109.1.1.1.1.2.3 = 5001
.1.3.6.1.4.1.9.9.109.1.1.1.1.3.1 = 10    #cpmCPUTotal5sec
.1.3.6.1.4.1.9.9.109.1.1.1.1.3.2 = 55
.1.3.6.1.4.1.9.9.109.1.1.1.1.3.3 = 0
.1.3.6.1.4.1.9.9.109.1.1.1.1.4.1 = 5     #cpmCPUTotal1min
.1.3.6.1.4.1.9.9.109.1.1.1.1.4.2 = 30
.1.3.6.1.4.1.9.9.109.1.1.1.1.4.3 = 0
.1.3.6.1.4.1.9.9.109.1.1.1.1.5.1 = 2     #cpmCPUTotal5min
.1.3.6.1.4.1.9.9.109.1.1.1.1.5.2 = 17
.1.3.6.1.4.1.9.9.109.1.1.1.1.5.3 = 0
	I've trimmed down the output for clarity.
	I can derive the index by REGEXing the OID .1.3.6.1.4.1.9.9.109.1.1.1.1.2, which allows me to
gather the 5 sec, 1 min and 5 min loads for each of the CPUs (three in this case).
	My problem is I also want to gather the "entPhysicalName" from a different OID
(.1.3.6.1.2.1.47.1.1.1.1.7) using the "cpmCPUTotalPhysicalIndex" reported by the first OIDs as an
index. A walk of them reveals:
.1.3.6.1.2.1.47.1.1.1.1.7.2017 = "CPU of Routing Processor"
.1.3.6.1.2.1.47.1.1.1.1.7.2001 = "CPU of Switching Processor"
.1.3.6.1.2.1.47.1.1.1.1.7.5001 = "CPU of Module 9"
(Continue reading)

Stephan Seitz | 20 Feb 10:05 2014
Picon

Questions about interface traffic statistics

Hi!

I’m using cacti 0.8.8a together with Debian 7 and have some questions 
concerning interface traffic statistics.

The poller runs every five minutes. The system has two gigabit network 
interfaces configured as bonding device. Via SNMP I’m polling bond0 
together with the two physical interfaces using „In/Out Bytes with Total 
Bandwidth”.

Now I notice that the numbers in cacti are much lower than compared with 
the ifconfig output. And the total values of bond0 are not the sum of the 
physical interfaces.
E.g. choosing a time range of 16 hours the cacti values are:
	- bond0: total in:  79GB, total out: 123GB
	- eth0:  total in:  36MB, total out: 167GB
	- eth1:  total in: 120GB, total out:  15MB

Calculating the traffic with the help of the ifconfig output I’ll get:
	- bond0: total in: 376GB, total out: 424GB
	– eth0:  total in:  34MB, total out: 424GB
	- eth1:  total in: 376GB, total out:  14MB

So something is wrong. There is the possibility to chose „In/Out Bytes 
(64bit counter)”, but then I don’t have the total bandwidth values. Could 
this be a counter problem?

Shade and sweet water!

	Stephan
(Continue reading)

Italo Valcy | 18 Feb 20:32 2014
Picon

CACTI authentication fallback to Local when LDAP fails


Dear all,

I think it would be nice if CACTI auth could fallback to Local realm
when LDAP fails to authenticate an user. For now the user has to
choose an auth method (realm) from login page. This seems little
inconvenient to a customer that just want to see his graphs and dont
worry about auth methods.

I've seem a bug report on this [1] and there said: "Fall through will
work and only work if LDAP does not return an error". However, looking
at the code (auth_login.php), in any way $ldap_error gets "true" value
and the Local fallback just dont work.

[1] http://bugs.cacti.net/print_bug_page.php?bug_id=1121

Is there any problems in having this feature? I can try to work on this.

--

-- 
Atenciosamente,

Italo Valcy
PoP-BA/RNP :: http://www.pop-ba.rnp.br
INOC: 53164*100
Tel.: +55 71 3283-5693

Bruno Martins | Grupo Rumos | 17 Feb 15:13 2014

Cacti not graphing mURLin data

Hi list,

We are facing a problem in which Cacti does not represent the data collected by the mURLin plugin.

We are using Cacti version 0.8.8b and mURLin version 0.2.2.

This is the output from /var/www/cacti/log/cacti.log:
02/17/2014 02:00:11 PM - CMDPHP: Poller[0] Host[60] DS[457] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[458] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[459] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[459] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[460] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[460] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[461] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[461] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[462] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:12 PM - CMDPHP: Poller[0] Host[60] DS[462] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:13 PM - CMDPHP: Poller[0] Host[60] DS[463] WARNING: Result from CMD not valid.  Partial
Result: U
02/17/2014 02:00:13 PM - CMDPHP: Poller[0] Host[60] DS[463] WARNING: Result from CMD not valid.  Partial
(Continue reading)


Gmane