Cacti and SNMP error
I've more than 10 servers(yeah yeah, I'm not here to compare dick sizes or buying habits) now, so I though having a self-hosted monitoring domain would be fine on my one of LES boxes. Based on this thread http://www.lowendtalk.com/discussion/54477/self-hosted-monitoring#latest, I've tried librenms, but due to my lack of understanding ditched it today and today I've tried cacti instead.
Cacti is throwing out snmp errors when adding a new device. I've used this tutorial on ubuntu 14.04 to setup the snmp agent http://www.unixmen.com/cacti-monitor-linux-servers-using-snmp/, then used http://download.dartware.com/docs/DevGuide/Content/02-CustomProbes/Debugging_with_SNMPWalk.htm for snmpwalk debugging and the following snmpwalk returns a bunch of values is which is expected from what I've read earlier. Code I'm using:
snmpwalk -v 2c -c communitystring1234random server.randomdomain.com
And it spits out a bunch of OID values as expected. I've tried both from the snmp manager(where cacti is hosted) and the snmp agent server and both receive the same response. Last line from response:
iso.22.214.171.124.126.96.36.199.0 = No more variables left in this MIB View (It is past the end of the MIB tree)
I believe I've read this meaning that the snmp agent sent over all information it has access to.
Cacti was installed from 0.8.8.f(latest tar.gz from cacti.net http://www.cacti.net/downloads/cacti-0.8.8f.tar.gz).
I've activated the following plugins:
Autom8 0.35 Automate Cacti Tasks General Active Reinhard Scheck Slowlog 1.3 Cacti MySQL Slow Log Viewer General Active The Cacti Group Aggregate 0.75 Create Aggregate Graphs General Active Reinhard Scheck Monitor 1.3 Device Monitoring General Active Jimmy Conner Settings 0.71 Global Plugin Settings System Active Jimmy Conner Nectar 0.35a Send Graphs via EMail General Active Reinhard Scheck Syslog 1.22 Syslog Monitoring General Active Jimmy Conner
The cacti host is running: nginx 1.9.5, php5.6-fpm dotdeb wheezy, mariadb 10.1, debian wheezy.
On new device addition fields I've got my snmp agent server as the following:
Hostname server.randomdomain.com Host Template Local Linunx Machine Number of Collection Threads 1 Monitor Host Ticked Downed Device Detection Ping and SNMP Uptime Ping Method ICMP Ping Ping Timeout Value 1000 Ping Retry Count 3 SNMP Version Version 2 SNMP Community communitystring1234random SNMP Port 161 SNMP Timeout 1000 Maximum OID's Per Get Request 10
The agent server's /etc/snmp/snmpd.conf file(only pasted the non-commented lines in here:
# AGENT BEHAVIOUR agentAddress udp:161,udp6:[::1]:161 # ACCESS CONTROL view systemonly included .188.8.131.52.2.1.1 view systemonly included .184.108.40.206.220.127.116.11 view all included .1 80 rocommunity communitystring1234random default -V systemonly rouser randomauthOnlyUser # SYSTEM INFORMATION sysLocation Provider, Provider Location sysContact [email protected] sysServices 72 # Process Monitoring proc mountd proc ntalkd 4 proc sendmail 10 1 # Disk Monitoring disk / 10000 disk /var 5% includeAllDisks 10% # System Load load 12 10 5 # ACTIVE MONITORING trapsink localhost public # Event MIB - automatically generate alerts iquerySecName internalUser rouser internalUser defaultMonitors yes linkUpDownNotifications yes # Arbitrary extension commands extend test1 /bin/echo Hello, world! extend-sh test2 echo Hello, world! ; echo Hi there ; exit 35 # AgentX Sub-agents master agentx
I don't understand what cacti's problem is. I've installed and double-checked all php extension needed by their documentation.
Ping between agent and manager is 8ms(London and Netherlands).
On devices tab, cacti has info on the local server, but detects my agent server to be down, which isn't true.