Started getting multiple events "Device status changed to Down (ping) " over the last few weeks with no major increase in number of devices.
Issue started on Cento 6.9 install with PHP 5.4, MySQL 5.1 (260 devices) HW - dual - quad core xeon CPUs, 24 GB Ram, RRD, MySQL, OS - with Observium install all on separate SSD drives Observium SVN install 17.6.8672 (rolling)
Upgraded PHP and MySQL to PHP70 with opcode caching enabled, MySQL 5.7.18 - no change
Spun up a fresh install Centos 7.3, PHP70 with opcode caching enabled, MySQL 5.7.18 (install per docs) HW - dual - quad core xeon CPUs, 24 GB Ram, RRD, MySQL, OS - with Observium install all on separate SSD drives Observium SVN install 17.6.8672 (rolling) $config['snmp']['max-rep'] = TRUE; Currently only 139 devices added
To eliminate DNS related issues I have added the devices to /etc/hosts and forced resolution to use it first. Cron has /opt/observium/observium-wrapper poller -w 5 running every 5 min.
The devices are not down and if I manually poll the device with a "Device status changed to Down (ping) " notice it responds normally and comes right back up.
Was wondering if anyone has seen any behavior like this with this build or before. Also is there a way to skip ping during the poller run and just do snmp, and if so is it something that I can enable globally?
Thanks in advance
Ron Culler