Ad blocker detected: Our website is made possible by displaying online advertisements to our visitors. Please consider supporting us by disabling your ad blocker on our website.
Please look at the attached graph and let me know if it seems correct. It does NOT to me, but I'm very new to Cacti. The port I am monitoring is the uplink from my server farm to the other buildings, so I would think the traffic would be A LOT higher. Am I right in assuming this shows only about 3-4 megabits per second? It's a gigabit fiber uplink from 7 servers to other buildings. Seems very low... I am using 64-bit counters.
You can re-check max values of your Data source. Cacti default value is 100.000.000 for both inbound and outboun. You must use 1.000.000.000 for gigabit interface.
Where have you said that? Rrdtool reads traffic counter via SNMP. This traffic counter counts bytes. Rrdtool therefore stores values in bytes/sec inside rrd file. For this reason you have to put maximum value also in bytes/sec. I know all of us display traffic graphs in bits/sec but it is only because we just use "Turn bytes into bits" CDEF just before plotting our data series. This has nothing to do with data collection and maximum value.
I know all of us display traffic graphs in bits/sec but it is only because we just use "Turn bytes into bits" CDEF just before plotting our data series.
So do I set the data output type ID to bytes per second (64-bit), use 125000000 as the max value, and then use the CDEF to translate it back in to bits?
What I want is a graph that shows the true value for bits per second transfered over the interface.
Remember that you are looking at 5 minute average values. We were shocket to discover how low our general throughput was. Infact, the busiest server here only manages to get above 200Mbps during backups. And this thing is connected with two gigabit cards!
So, in your graph you are averaging close to 288 Megabytes of data transferred every 5 minutes! I don't know what you have attached on this port, but you are showing a fair amount of data.
So, in your graph you are averaging close to 288 Megabytes of data transferred every 5 minutes! I don't know what you have attached on this port, but you are showing a fair amount of data.
I have 7 servers at this site, and that port is the uplink to the main gateway. That's why I was wondering why it was so low. Didn't think about the 5 minute factor.
So do I set the data output type ID to bytes per second (64-bit), use 125000000 as the max value, and then use the CDEF to translate it back in to bits?
I just mean that regardless of your graph settings maximum value should be set in DS settings in bytes/sec (for normal 32/64bit SNMP traffic counters).
That's why I was wondering why it was so low. Didn't think about the 5 minute factor.
5 minutes sampling period will make the graph looking more "flat". If you think the server is loaded with traffic for longer period of times (like 30 min of backup process) then you should see high utilization on the link anyway. If you would like to see the traffic in real-time then use this simple tool: http://leonidvm.chat.ru/ .
I had a similar shock. We've got a 10gig link between our datacenters that's never seen over 220mbits/sec, and that's on a 1 minute average. On a 5 minute average I don't think I ever saw it top 100mbits/sec.
