eG Help

Sometimes, a server may be functional and connected to the network, but the network connectivity may not be good enough to provide good throughput. This may result in applications hosted on the server offering poor response to users. Speed mismatch between the network interface card on the server and the switch port that it is connected to is often one of the main reasons why such throughput issues occur in production environments.

The ThroughputTest aims to monitor and detect situations when the throughput for data transfer to and from a server drops below acceptable levels. This test is executed by the eG agent installed on a server, and it emulates an upload and a download action from a server. Upload involves data transfer from the agent on the server, while download involves downloading of content to the server from a remote location. The upload and download activities are initiated against the eG management console to which the agent reports. Hence, the throughput metrics report the values that users can expect when transmitting data between the server being monitored and the eG management console.

Run by the internal agent. HTTP is the protocol used for the throughput measurements.

Measurement	Description	Measurement Unit	Interpretation
Upload_availability	This metric indicates whether the upload data transfer succeeded or not. This metric takes a value of 0 if the data upload to the eG management console fails. A value of 100 indicates that the upload completed successfully.	Percent	Upload failures indicate either a problem with the eG management console, or a failure of the network routing to the eG management console (e.g., because one of the routers on the path to the eG management console is down).
Upload_speed	This metric represents the throughput seen during upload transfers (i.e., from the server being monitored to the eG management console).	Kbps	A significant reduction in this value over time indicates a potential problem scenario that needs investigation - i.e., is it the network or the eG management server that is causing the slowdown (if the bottleneck is at the eG management server, the slowdown would be seen across the board by all the agents).
Download_availability	This metric indicates whether the download data transfer succeeded or not. This metric takes a value of 0 if the data download from the eG management console fails. A value of 100 indicates that the download completed successfully.	Percent	Download failures indicate either a problem with the eG management console, or a failure of the network routing to the eG management console (e.g., because one of the routers on the path to the eG management console is down). It is especially important to monitor upload and download throughput values for networks that have different downstream and upstream characteristics like ADSL or Hybrid Fiber Coaxial networks.
Download_speed	This metric represents the throughput seen during download transfers (i.e., from the eG management console to the server being monitored).	Kpbs	A significant reduction in this value over time indicates a potential problem scenario that needs investigation - i.e., is it the network or the eG management server that is causing the slowdown (if the bottleneck is at the eG management server, the slowdown would be seen across the board by all the agents).