I have a question about the roles of the following settings for Jrun Web Server and Jrun Proxy Server:

ActiveHandlerThreads
MaxHandlerThreads
Backlog

My understanding is that if I have the following settings in jrun.xml:
    <attribute name="activeHandlerThreads">200</attribute>
    <attribute name="backlog">500</attribute>
    <attribute name="maxHandlerThreads">1000</attribute>
    <attribute name="minHandlerThreads">200</attribute>
... then up to 200 threads will run concurrently. If the web site traffic results in more than 200 concurrent requests, then additional threads are spawned and will be assigned to request 201 onwards. These extra threads remain idle until one of the active threads completes.

>From what I can see then, making maxHandlerThreads greater than activeHandlerThreads is effectively creating a queing mechanism. So, my settings (above) will allow for a queue of up to 800 idle threads (1000-200).

First question:
Am I understanding this correctly?

Second question:
What is the difference between maxHandlerThreads and the �backlog� setting? Why do we need the �backlog� setting when we can effectively create a queue by making maxHandlerThreads greater than activeHandlerThreads.

I�d be really grateful for any help on this.

Thanking you,

Gavin
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings]

Reply via email to