Hi All,
I’ve implemented a distributed poller that is working 100%
I noticed yesterday that one of my devices stoped graphing I am able to poll from the new poller.
after I changed this device back to the default poller it started graphing.
validate.php shows all is good on both pollers.
my second poller is registered as a node aswell.
but like I said 80 of my devices are graphing fine just this one device stops graphing.
if you guys maybe have some troubleshooting tips I can use please
This is how it looks found another host.
running a manual poll.
./poller.php 172.xx.xx.xx 2020-03-18 10:19:51 - 1 devices polled in 24.51 secs
SNMP [4/1.02s]: Get[4/1.02s] Getnext[0/0.00s] Walk[0/0.00s]
MySQL [32/11.73s]: Cell[17/5.70s] Row[-17/-5.70s] Rows[20/6.72s] Column[0/0.00s] Update[10/4.21s] Insert[2/0.80s] Delete[0/0.00s]
RRD [6/0.00s]: Update[6/0.00s]
Are both pointing at the same rrdcached server? (you should only have one)
I have one rrdcached on my Default Setup and my secondary is pointing to it in config.php
Maybe a tcpdump to verify the rrd updates are coming in? Try to figure out where the issue is.
Tested that aswell I can see Established connection from Secondary to Primary on the specified port.
My issue is that I have 80 hosts on my secondary and 78 of them are 100% (the hosts I’ve noticed) 2 of them are not graphing its really weird bro
This is the RRD error
librenms.log
[2020-03-20 08:46:45] production.ERROR: Call to undefined function rrdtool_build_command() {“userId”:2,“exception”:"[object] (Symfony\Component\Debug\Exception\FatalThrowableError(code: 0): Call to undefined function rrdtool_build_command() at /opt/librenms/includes/html/graphs/graph.inc.php:120)
is the following config needed in my second poller?
##RRD
$config['rrd_dir'] = '/mnt/data-01/rrd';
probably yes
Update your install to fix that.
On both my servers
rrdtool is already the newest version (1.7.0-1build1).
Update your LibreNMS install.
Both my Installs are running the latest version of LibreNMS…
Could the issue be in my cronjobs?