I have run into this a few times since we opened up a few years ago. Small business networks that have intermittent network packet loss from computer to computer, sometimes also comes with internet packet loss.
Generally this has been caused by bad hardware. Two examples that come to mind: one was a bad NIC and one was a bad KVM switch that included Ethernet.
In one of these cases they would completely lose network connectivity on all systems. Power cycling the switch fixed it for 24-72hrs. Replaced the switch, but it didn't fix the problem. Ended up being a faulty KVM switch that had a NIC port being used.
The second case all systems on the network would randomly drop about 12-18 packets a day. The dropped packets would happen at different times during the day. For example system A would drop some packets in the AM, system B would drop some in the PM. It might be the exact opposite the next day. No system was worse or better than any other system. It would happen to all of them. Ended up being a faulty NIC in one of the systems.
Basically, the only way I knew to work on these problems was to use a constant ping -t and gauge the results. Segment the network until we eventually found the culprit causing the issue.
This is very time consuming and not efficient.
Does anyone have any suggestions on how to better handle these situations?
I appreciate you taking the time to help.
Generally this has been caused by bad hardware. Two examples that come to mind: one was a bad NIC and one was a bad KVM switch that included Ethernet.
In one of these cases they would completely lose network connectivity on all systems. Power cycling the switch fixed it for 24-72hrs. Replaced the switch, but it didn't fix the problem. Ended up being a faulty KVM switch that had a NIC port being used.
The second case all systems on the network would randomly drop about 12-18 packets a day. The dropped packets would happen at different times during the day. For example system A would drop some packets in the AM, system B would drop some in the PM. It might be the exact opposite the next day. No system was worse or better than any other system. It would happen to all of them. Ended up being a faulty NIC in one of the systems.
Basically, the only way I knew to work on these problems was to use a constant ping -t and gauge the results. Segment the network until we eventually found the culprit causing the issue.
This is very time consuming and not efficient.
Does anyone have any suggestions on how to better handle these situations?
I appreciate you taking the time to help.