Hello,
Wanted to use Zabbix network discovery in env with an autoscaling group with lots of hosts, the issue is I have /21 network and it took forever for 1 discoverer to scan it. I've tried to split rules to /24 ranges, but still, it takes about 15 minutes to scan it all.
And I want it to be faster, so if the new instances spawn they will be registered and monitored as fast as possible. Any way to do it?
Also, what is the usual approach to undiscovery in such cases? Cause I won't host to alert in case it freezes or is unavailable in some way, but didn't alert if it was terminated by autoscaling rules.
Thanks!
Wanted to use Zabbix network discovery in env with an autoscaling group with lots of hosts, the issue is I have /21 network and it took forever for 1 discoverer to scan it. I've tried to split rules to /24 ranges, but still, it takes about 15 minutes to scan it all.
And I want it to be faster, so if the new instances spawn they will be registered and monitored as fast as possible. Any way to do it?
Also, what is the usual approach to undiscovery in such cases? Cause I won't host to alert in case it freezes or is unavailable in some way, but didn't alert if it was terminated by autoscaling rules.
Thanks!
Comment