Ask any network guru what they do first when taking on a trouble ticket, and inevitably they’ll say “PING!” What is PING? By sending ICMP (Internet Control Messaging Protocol) Echo Request frames and listening for the Echo Reply frames, PING can give us assurance ofsome level of connectivity, response time, and packet loss.
ICMP over IP gives us an “inside out” starting point for troubleshooting… if the target host replies, we KNOW we have a path (connectivity) and a “rough idea” of network round trip time. If a host does not reply, well… that can mean that we don’t have a path (route), the device is offline, it was too busy to respond, the ping packet got discarded, ICMP could be blocked…and so on. So an unsuccessful PING just means, “time to look somewhere else”.
Response times in PING are only useful if you know what the NORMAL response time is to/from that network and host – but only the host’s NIC, as ICMP echo requests are handled by lower layers. Measuring packet loss via ping could be good information generally, but is still unreliable given that ICMP is often discarded by infrastructure devices. A much more reliable method is to utilize a TCP Port Open command. Ping is not performance!
*PS – Amaze your friends with this handy bit of trivia… What’s the acronym PING stand for? Packet Inter Net Groper. Well, maybe… read the story HERE.