I am having an ongoing problem with my colocation provider. They generate
graphs via RDD tool and use the data to bill their customers. I am looking for
a definitive answer as to what can be happening.
Some additional info / background. I have been their customer for over 5
years. About once or twice a year for a period of 24 hours, our graphs max
out. The time the graphs max out coincides with us moving backups from one
server to another. We do this weekly. Why the internal traffic is being
caught by the router is a another topic. The info is only provided so that you
have all the info regarding the problem. We commit to 3 mbps and are capped at
10 mbps and are connected to a 100 mbps ethernet interface.
Up until a couple of years ago, when the problem occured, our rrd graphs would
spike to 100 mbps. Since we were capped at 10 mbps I would call my customer
service and since the traffic was over our capped rate, they would credit me
the overage charges.
A few years back, I mrtg to monitor the traffic on all four of the servers I
have colocated. Every time this event happens, it is at the time we transfer
data from one server to another on the internal network. My mrtg graphs
clearly show high inbound traffic on one server, and high outbound traffic on
the other.
A couple of years ago, they said the "fixed" the problem. The next time this
event happened, my mrtg graphs showed exactly the same thing. 5 minutes of 100
mbps traffic from on server to the other. My colocation providers graphs
showed something totally different. They showed 10 mbps traffic for nearly a
day.
Now I have read rrds docs and tutorials. They all state that any value over
the configured maximum is ignored. My question is how is it possible for my
graphs to be showing 5-10 mins of 100 mbps traffic and my colocation providers
graphs show almost 24 hours of 10 mbps traffic?
Thank you in advance,
John Maslanik
_______________________________________________
rrd-users mailing list
[email protected]
https://lists.oetiker.ch/cgi-bin/listinfo/rrd-users
