NeophyteI have a SQL Server Listener (part of AG). It has Ping critical alert showing up and clearing. It is happening every 20 or 30 minutes. I attach the screenshot.
It clears and it's good for a while and then it comes back again. Would this indicate the network issues?
Another example of Ping on different Listener is that is shows Warning for Ping and also shows up and clears multiple times a day.
ProfessorWhat is the ping alert reporting? 100% packet loss or something less? or high average round-trip time? Can you replicate it outside of LogicMonitor? Like running ping -t from the collector server or perhaps another 3rd server? If the collector server is not on the same subnet, do you see ping issues when on the same subnet?
Unless the target is running on super old hardware, there should be a TCP Offload Engine (TOE) on the NIC that should be handling things like ping responses without consulting the kernel: meaning that it's not likely due to the target.
Yes, I'd suspect the network first. In addition to Mike's questions: are there any firewalls (adaptive/dynamic ones) between the collector and the server? Is it possible you have an IP address conflict on the network?
NeophyteMost likely issue is the ping not resolving to the right IP address. 2 devices in Logic Monitor are added as Listeners- they are both multi-subnet Listeners- ping is trying to hit one of the secondary IPs instead of online listener IP. So it just fails until TTL expires and it retries and we get lucky it starts to ping the right IP address until TTL expires and retries again.
ProfessorYou can change the device in LM to use IP addresses instead of fqdn in the "IP Address/DNS name" field.
I'm assuming "mutli-subnet listeners" means the server is multi-home and has multiple nics/ips which allow it to connect to different subnets without routing.
NeophyteSo the problem is that I can't assign one IP because every time there is a failover it will switch between those 2 IP addresses and if there's just one IP assigned to device in Logic Monitor and it will pick other IP from the subnet it will recognize it as incorrect. FQDN of listener resolved Host Status issue but didn't resolve the Ping issue.
