greetings.
I am trying to put some figures together for an ROI.
I landed a couple Sun Netra T1s, 440Mhz, 512MB RAM, 1X 18GB disk (10K RPM). The hardware is locked, so I am stuck using it.
I am going to impliment a squid reverse proxy on these servers, and could use some help from anyone who pushes a decent amount of traffic through squid.
I have the following questions.
What are you currently pushing from Squid?
Max Connections / sec
Average connections / sec
Average byte size
What is your hardware config?
What is your squid config?
The squid server I am setting up is acting as a reverse proxy.
Assuming you were implimenting squid on this platform for this function, and you couldn't change the hardware, and you needed the server to perform exceptionally fast; from this performance perspective, what could one expect from Squid (Max Connections a Second).
I know this question is vegue without providing details on file size, etc. Best guesses would be helpfull.
Clustering Squid. I read that squid has the ability to cluster.. I.E. if one of my two squid servers gets a request, but does not have a copy of the requested file, before going back to the source, it can check the other cache servers in the cluster to see if another server has the file.
Is this true, and if so, what is the performanc hit? How does this work?
What is the ratio for memory and disk for optimal performance.
I have a zillion more questions. Any help would be appreciated.
RT
Received on Sun Oct 21 2001 - 18:46:23 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:03:01 MST