מאמר שנכתב ע"י Richard Steenbergen הוא הקים את nLayer.
הוא מסביר בפירוט על Traceroute
אני למדתי המון מזה, ממליץ בחום.
המאמר המלא:
http://cluepon.net/ras/traceroute.pdf
פסקה מהמאמר:
[left]
Queuing Delay
To understand queuing delays, first you must understand the nature of interface utilization. For
example, a 1GE port may be said to be “50% utilized” when doing 500Mbps, but what this actually
means is 50% utilized over some period of time (for example, over 1 second). At any given instant,
an interface can only be either transmitting (100% utilized), or not transmitting (0% utilized).
When a packet is routed to an interface but that interface is currently in use, the packet must be
queued. Queuing is a natural function of all routers, and normally contributes very little latency to
the overall forwarding process. But as an interface approaches the point of saturation, the
percentage of packets which must be queued for significant periods of time increases exponentially.
The amount of queuing delay that can be caused by a congested interface depends on the type of
router. A large carrier-class router typically has a significant amount of packet buffering, and can add
many hundreds or thousands of milliseconds of latency when routing over a congested interface. In
comparison, many enterprise-class devices (such as Layer 3 switches) typically have very small
buffers, and may simply drop packets when congested, without ever causing a significant increase inlatency.
[/left]
