Notice: This product has been discontinued as of 20th March 2019 from our product list. Any further development and all related support services would also be discontinued. You can, however, continue to use this product, and any related support services would come under our per hour administrative charge of $25 (USD).
The Admin-Ahead Arch for Apache (Plesk) allows web hosts to gain in-depth knowledge about the performance of their Apache web servers and plan future hardware and software scalability requirements, as well as find bottlenecks and unusual activity that may be hindering the smooth running of Apache.
Below you can find a brief description of each graph.
Total Access:
This graph shows the total number of access to the server. But the counter will get reset to zero every time the server is restarted. This is an ever-increasing value ( until the server is restarted ). This graph can be used to analyze the rate at which the number of access is increasing.
Total Traffic (MB):
This graph shows the total traffic handled by the server (in MegaBytes ). This can be very handy at times. Analyzing this graph helps us to identify the times at which most traffic was present. For example, if the server is running some media websites, we can identify the traffic patterns and thus enable measures to balance the load at peak times. The server won't be receiving the same traffic all day, identifying the traffic pattern is really important so that the server does not get crushed under heavy load unexpectedly.
Requests Per Second:
It means the number of requests handled by the server in a second. Sometimes the server gets high traffic and the CPU usage goes up. By analyzing the graph, we can predict the possible number of requests per second in the future
Bytes Per Second:
This graph shows the total bytes served in a second. This is important because the requests per second parameter just find out the number of requests alone, it could be a request to a simple static page or a heavy dynamic page. But, bytes per second helps to identify the times during which most contents were served
Bytes Per Request (K):
This shows the number of bytes served for each request. This is calculated by dividing the Bytes Per Second by Requests per second.
Busy Workers:
In prefork MPM ( Multiprocessing module ), the number of workers means the number of apache processes spawned. Busy workers mean the total number of apache processes that are handling requests by the clients. And, in worker MPM the number of workers means the number of threads created by apache. As the name suggests, they are busy handling requests. If the number of busy workers is high, it means the server is receiving heavy traffic.
Idle Workers:
By watching the number of busy and idle workers, we can find the problems in our Apache configuration. If the number of idle workers reaches zero during a peak time, then such queued requests must wait for older requests to be processed. This could result in slower response times for the website.
To deal with the situation, we can increase "MaxClients" which is the limit for the number of simultaneous connections. But, this should be handled with caution, if the server does not have enough ram to spare for the extra workers, we should not increase the "MaxClients". If we increase the MaxClients to a high value, and the server runs out of ram, it will start to use the swap space and it will drastically affect the performance. We should never let the server touch the swap space by any means. If the RAM is getting filled, lower the "MaxClients".