Back to Tools

Ping Test — Network Reachability Checker

Test host reachability, measure latency, and detect packet loss from the OpsCheck server.

Enter a hostname or IPv4/IPv6 address (e.g., google.com or 8.8.8.8)
Ping Results for

ICMP Protocol: The Foundation of Network Diagnostics

Ping uses the Internet Control Message Protocol (ICMP), a network-layer protocol that operates alongside IP. Every sysadmin uses ping daily, but few understand the protocol details that explain false positives and false negatives.

ICMP TypeNameWhat It Means
Type 0Echo ReplyThe target received your ping and is responding. This is what you want to see.
Type 8Echo RequestYour machine sends this. Contains a payload and a sequence number for matching replies.
Type 3Destination UnreachableA router in the path cannot reach the target. Code field tells you why: network (0), host (1), protocol (2), port (3), fragmentation needed (4).
Type 11Time ExceededTTL hit zero before reaching the target. This is how traceroute works — sending probes with increasing TTL values.
Type 4Source QuenchDeprecated. Originally meant to ask the sender to slow down. Modern networks use TCP congestion control instead.

How a Ping Packet Travels: Step by Step

  1. Application layer: The ping utility constructs an ICMP Echo Request (Type 8) with a payload (typically 32-64 bytes of timestamp and padding data) and a TTL (usually 64 on Linux, 128 on Windows).
  2. Raw socket: Ping opens a raw socket (requires root/Administrator privileges or CAP_NET_RAW on Linux). Unlike TCP/UDP, ICMP has no port numbers — it is identified purely by type and code fields.
  3. Routing: The kernel looks up the destination in its routing table. The packet is sent to the next-hop gateway or directly to the destination if on the same subnet.
  4. Response: If the destination receives the Echo Request, its IP stack generates an Echo Reply (Type 0), swaps source and destination addresses, copies the payload, and sends it back.
  5. RTT calculation: When the reply arrives, ping subtracts the sent timestamp from the current time to compute Round-Trip Time. Multiple pings reveal jitter (RTT variance), which is often more diagnostic than a single RTT value.
Key fact: A failed ping does not necessarily mean the host is down. Many firewalls, cloud providers, and ISPs block ICMP while allowing TCP traffic. A "Request Timed Out" can mean the host is up but ICMP is filtered, or the host is genuinely unreachable.

3 Common Ping Troubleshooting Scenarios

100% Packet Loss — Complete Silence

  1. Firewall blocking ICMP — many cloud providers (AWS, Azure) block ICMP by default; the host may be running fine on TCP
  2. Verify with an alternative: try nc -zv host 443 or curl -sI https://host to confirm reachability over TCP
  3. Check if ICMP is blocked at your side — corporate firewalls and VPNs often filter outbound ICMP
  4. For IPv6 targets, use ping6 or ping -6 — IPv4 ping to an IPv6-only host will always fail

High and Inconsistent Latency (Jitter)

  1. Check the first hop — high latency to your own gateway indicates local network congestion or Wi-Fi interference
  2. Run a traceroute/MTR to see which intermediate hop introduces the latency spike
  3. High jitter (RTT variation >20ms) impacts VoIP, gaming, and video streaming more than raw latency
  4. Geographic distance: light in fiber travels ~5 microseconds per km. Transatlantic RTT (~100ms) is mostly physics

Intermittent Packet Loss (Drops)

  1. Run an extended ping: ping -c 100 host — occasional drops (<1%) may be normal, sustained drops (>5%) indicate a problem
  2. Check for MTU issues: ping -M do -s 1472 host tests if packets fragment correctly across the path
  3. ICMP rate limiting: many routers limit ICMP responses to 100-1000 packets/second. Your target may simply be throttling replies
  4. ECMP (Equal-Cost Multi-Path) can cause asymmetric routing — some paths may have issues while others work fine

5 Ping Gotchas That Mislead Sysadmins

1. ICMP deprioritization: Many ISPs and routers treat ICMP as low-priority traffic. During congestion, ping packets are dropped first while TCP traffic flows normally. High ping loss during peak hours may not indicate a real problem.
2. MTU black holes: If a router on the path cannot fragment a packet (DF bit set) and the packet exceeds the link MTU, the router should send ICMP Type 3 Code 4 (Fragmentation Needed). Some routers silently drop instead. Use ping -M do -s 1472 to detect these.
3. TTL expired vs unreachable: "TTL expired in transit" means a routing loop or an exceptionally long path (>64 hops). "Destination Host Unreachable" means an ARP failure on the last-hop router. These are different problems with different fixes.
4. IPv6 ICMPv6 blocking: IPv6 relies on ICMPv6 for neighbor discovery, SLAAC, and Path MTU Discovery. Blocking ICMPv6 completely breaks IPv6 — unlike IPv4 where blocking ICMP is mostly safe. Filter selectively, not wholesale.
5. Cloud load balancer ping: When you ping a cloud load balancer's IP, you are pinging the load balancer infrastructure, not your actual server. The load balancer may respond while your server is down. Always check your application endpoint, not just ICMP.

CLI Equivalent

# Standard ping (4 packets)
ping -c 4 example.com
# Extended ping with specific packet size
ping -c 10 -s 1400 example.com
# Continuous ping (Ctrl+C to stop)
ping example.com
# IPv6 ping
ping6 -c 4 example.com

Related Tools

Port Scanner DNS Lookup HTTP Headers

It means your ping did not receive an Echo Reply within the timeout period (typically 1-2 seconds). This can indicate: the target host is down, a firewall is blocking ICMP, the network path is broken, or the target is simply configured not to respond to pings. Many cloud servers have ICMP blocked by default.

<20ms is excellent (same city/datacenter). 20-50ms is good (same country). 50-100ms is acceptable (same continent). 100-200ms is slow (intercontinental). >200ms indicates significant delay. Jitter (variance between consecutive pings) matters more for real-time applications — <5ms jitter is ideal.

Many organizations intentionally block ICMP at their firewall or load balancer to reduce attack surface (ICMP flood DDoS, ping sweep reconnaissance). The website can be fully functional while ping returns no response. Use a port scanner or HTTP check to verify the service is actually up.

Ping measures end-to-end reachability and round-trip time. Traceroute shows each router hop along the path by sending packets with incrementing TTL values. Use ping for a quick health check; use traceroute/MTR to identify where exactly packet loss or latency occurs.