Can anyone explain the correlation between response time and circuit utilization? We have put a 20% load on a test network and see response times climb from 18ms to around 190ms. Is this normal? If not, what should it be? Is there a formula of some sort that can show us what to expect on a circuit?
The formulas you are looking for are in the field of queueing theory. You should expect average delay to double at 50% of load, quadruple at 75% loading, and up by a factor of ten at 90% load. If your change in response time is due to queueing delays, then something in your network is more than 90% loaded, just not the link you were looking at. On the other hand, it is also possible, though less likely, that there are other problems, such as dropped packets, which are causing the performance degradation.
Network delay, especially over a WAN link that is small, can vary considerably depending on how it is tested.
When your network is unloaded your ping is fired out instantly from a router onto a small link (say 64kbps). However even when you lightly load your WAN link your ping might be forced to wait behind a 1500byte packet before it gets a chance to make it's way across the link.
So what's the big deal? Answer: serialisation delay.
How long does a 1500 byte packet take to transmit out onto a 64Kbps link?
Formula: time = packet_size / link_speed
So, time = 1500 bytes / 64 Kbps
..time = 1500 * 8 bits / 64,000 bits
..time = 187.5ms
There's your answer, and it has nothing to do with propagation delay. It's just queueing. Your ping has to wait behind other packets before it even gets a chance to encounter propagation delay on the WAN link. That's even with round-robin or priority queueing!!
How to get it down? Force the MTU to something low maybe (not a good answer), or buy a bigger bandwidth link!
Note that 1500 byte packets will be common because that's the maximum size an ethernet packet may be (approx). And when doing file transfers or HTTP transfers the packet sizes will be as big as possible.
I suspect your WAN link may actually be 128Kb/s in size because the serialisation delay quoted above was one-way.. and a ping is always a round-trip measurement so two serialisation delays will always be encountered (minimum).. of course this increments to 3 or more if you traverse an ATM network on the way..
Question We run asr9001 with XR 6.1.3, and we have a very long delay to
login w/ SSH 1 or 2 to the device compare to IOS device. After
investigation, the there is 1s delay between the client KEXDH_INIT and
the server (XR) KEXDH_REPLY. After debug ssh serv...
Introduction The purpose of this document is to demonstrate the Open
Shortest Path First (OSPF) behavior when the V-bit (Virtual-link bit) is
present in a non-backbone area. The V-bit is signaled in Type-1 LSA only
if the router is the endpoint of one or ...
Hi, I am seeing quite a few issues with patch install and wanted to
share my experience and workaround to this. Login to admin via CLI, then
access root with the “shell” command Issue “df –h” and you’ll probably
see the following directory full or nearly ...