Network Engineer at a government with 1,001-5,000 employees
Real User
2021-02-04T12:01:36Z
Feb 4, 2021
Only for bandwidth, you can use MRTG. You can also copy all the trafic to a probe and use NTOPNG to have a dashboard with also the protocols flows and other IP data.
Search for a product comparison in Network Monitoring Software
One I am looking closely at is AppNeta. They have an appliance that can digest the flow and do a better job than Netflow.
The other one we are using is ExtraHop. This has both a Datacenter High Volume version and also an Appliance one could use at a remote Facility. We are using the smaller appliances at our Hospitals and their Uplink to the WAN Routers. Good job on showing the 1 second burst values as well as longer term usage.
It also can help with studies of HTML / HTTP applications. I put a simple trigger in place to report on transactions that finish in different 'buckets' 1 - <3 seconds; 2 - >3 and < 10; 3 - >10 and < 60; 4 - over 60. We are looking to establish a SLA of 99.9% under 3 seconds. This can report down to the 5 minute interval size and has spotted spikes when the back-ends are doing odd things. Those 5 minute time frames are when the end users are having pain with the application.
Of course when they call it in to the Help Desk - the problem has resolved. Now we have evidence that it did happen. Now we can ask them to fix it.
If they do not meet that SLA - $$$ can be lost from the standard fee because of failure to meet our requirements.
As expected they are balking. But - our end users are happy that now there is proof there have been issues. They are pressing to either solve the issues or find another vendor.
In Open Source - anything that can read the SNMP MIB of the Interface will work. But it does not tell you why it is being used - - Considered my other comment.
Only for bandwidth, you can use MRTG. You can also copy all the trafic to a probe and use NTOPNG to have a dashboard with also the protocols flows and other IP data.
PRTG is a nice tool to monitor performances of a wide kind of network devices. This tool is based on SNMP requests, and it is very easy of use
One I am looking closely at is AppNeta.
They have an appliance that can digest the flow and do a better job than Netflow.
The other one we are using is ExtraHop. This has both a Datacenter High Volume version and also an Appliance one could use at a remote Facility. We are using the smaller appliances at our Hospitals and their Uplink to the WAN Routers. Good job on showing the 1 second burst values as well as longer term usage.
It also can help with studies of HTML / HTTP applications. I put a simple trigger in place to report on transactions that finish in different 'buckets' 1 - <3 seconds; 2 - >3 and < 10; 3 - >10 and < 60; 4 - over 60.
We are looking to establish a SLA of 99.9% under 3 seconds. This can report down to the 5 minute interval size and has spotted spikes when the back-ends are doing odd things. Those 5 minute time frames are when the end users are having pain with the application.
Of course when they call it in to the Help Desk - the problem has resolved. Now we have evidence that it did happen. Now we can ask them to fix it.
If they do not meet that SLA - $$$ can be lost from the standard fee because of failure to meet our requirements.
As expected they are balking. But - our end users are happy that now there is proof there have been issues. They are pressing to either solve the issues or find another vendor.
Cacti is a simple open source too that focuses on collecting stats and graphing them.
https://www.cacti.net/
Open NMS is a more full featured monitoring platform with many more tools.
https://www.opennms.com/about/
In Open Source - anything that can read the SNMP MIB of the Interface will work.
But it does not tell you why it is being used - - Considered my other comment.