Why would receive bandwidth on a port show above what it is capable of doing?
Hi Debbi,
i think you are collecting those statistic data every 10 minutes right?The blue line shows the "real" data you have collected while polling.The bars show you a suggested min/max data that could also be theoreticaly be reached during the period that you are not polling that device.
If you collect those statistic data more often those min/max data will be more accurate.
What's interesting in this case is, how often Orion collects data while polling a deviceonce. For example: if i do a icmp ping on a device using the good old fashion Dos-Box i'll get 4 replys by default. What will Orion do in that case. Will he ping the device once or will orion ping the device more times and calculate the average response time? Same with those Statistic Data. How often will he collect the Data during one Poll to calculate the average receive utilization?
Maybe someone here knows this.
greetingsstefan
I have same kind of question regarding a Point to Point DS3 link... My polling is set to 5 minutes and I will occasionally see min/max reach up to 60 Mbps..
Maybe somebody in Solarwinds can explain why this is occuring...
My understanding was that the blue line was the average and the bars were the actual! -Debbi
The blue line is the average. The bars represent the range of data. If you're looking at 1-hour increments and you're polling every 5 minutes, then you'd get 12 samples. From those 12 samples, you would compute an average for the hour. You would then take the largest and smallest value of those 12 and that would define the two ends of the bar.
we understand that we would get 12 samples in a 60 minute period but the question is why are we seeing samples of 60Mbytes on a 45Mbyte link. This is physically not possible on serial links.
And in my case on a 100 Mbps port. I searched pretty thoroughly in past forum threads and found a few where this issue was brought up, but no one at SW ever really offered an explanation. The thread just disappeared into the past. I have another link where I saw the same thing this last week.
Debbi
SO .... I bring up the point again to Solarwinds.. Can you explain why we are seeing this? (i.e. DS3 Link with occasional Max of 60 Mbps)..
Please respond
Hi,
This type of error has historically been because of some misinformation from the device. We have seen slight differences in the maximum when the device is reporting information properly to show that the interface might be showing up to around 105% utilization. This is possibly due to the way the device reports the number of octets that it has sent. If it reports an entire packet's worth of bytes before it actually sends the packet, then this might account for a slightly off representation.
This calculation in NPM is truly one of the most simple where we calculate the delta of octets between the two polling intervals and divide that by the time between polls. If you are seeing this on a device consistently, use a MIB walk like tool, then you can poll the octets OIDs yourself, wait 20 seconds, poll again and do the calculation. NPM is a slave to the information we receive from the device so making sure the device is sending the right data is the first step.
We have seen issues with many types of devices where a firmware upgrade solved the problem. If the issue is easily reproducible or consistent, then it would be helpful to open a ticket with support so we can take a closer look at your diagnostics and device.
Thanks
