Handling problems monitoring low latency networks

Low latency networks are the ideal media; in typical intranet all the hosts can be reached in a few milliseconds.

Monitoring tools, starting from typical default ping utility (the one that sends ICMP Echo packets and waits for response), are consistent in their results if typical response is above 10-15 milliseconds. When response time drops significantly and is circa 1ms or even less, the results may begin to vary considerably.

IPHost typical response time when visible difference in response time measurement can be detected is circa 10ms. Everything above is usually consistent with results of the mentioned ping utility. However, if you are dealing with low latency networks and it is significant to have response time matching those reported by standard ping utility, you might need to replace your default PING monitors.

Using wrappers around standard PING utility

IPHost is supplied with bundled Python 3 interpreter, suitable for most custom monitoring tasks. Using this sample script (archive size 503 bytes) one can execute standard Windows ping utility and extract performance data (average ping response) from its summary. The script requires host domain name ($HostDNS template variable) or IP address as parameter.

Whereas that adds some insignificant delay to monitor polling time, it can provide users with the same results visible from default Windows ping tool.

You can also consider using alternate ping utilities.

PsPing from SysInternals

Well-known SysInternals collection of tools includes PsPing utility, suitable as standard ping utility replacement.

PSPing allows using TCP to probe a remote device; thus it can also be used as IPHost built-in TCP monitor replacement (to test low latency TCP-based services). Using the same wrapper approach, you can alter the mentioned above Python script to call PsPing and provide the stats as performance value.

fping tool

A general purpose fping utility can be named a Swiss knife of all ping tools; it provides low-level controls over sending and receiving ICMP Echo, both over IPv4 and newer IPv6 networks.

fping is provided in source code; if necessary, it can be used either from WSL (Windows Subsystem for Linux) internal virtual Linux environment, or from Posix-type simulators like Cygwin. The latter is quicker to integrate as IPHost monitors. You can consider fping as still another replacement for standard ping utility, and use it with corresponding wrapper script.

Conclusion

If you need assistance implementing any of the pinging approaches mentioned above, or if you have any related idea to offer, feel free to contact us or just leave a comment below.