• Introducing WebPros Cloud - a fully managed infrastructure platform purpose-built to simplify the deployment of WebPros products !  WebPros Cloud enables you to easily deliver WebPros solutions — without the complexity of managing the infrastructure.
    Join the pilot program today!
  • Support for BIND DNS has been removed from Plesk for Windows due to security and maintenance risks.
    If a Plesk for Windows server is still using BIND, the upgrade to Plesk Obsidian 18.0.70 will be unavailable until the administrator switches the DNS server to Microsoft DNS.

Sitebuilder 4.0 server capacity planning

T

terrychapman

Guest
Hello,

Does SWsoft(or any customers reading the forum) have any benchmark data regarding the typical user capacity for Sitebuilder on a reasonably spec'd modern server?
I realise that the answer could depend on a number of factors, however I am after a ball park figure based on averagely heavy Sitebuilder usage only on a typical modern dual processor, 2GB RAM, RAID 5 server.

Thanks in advance.

Terry
 
Hi Terry:

Where you able to get answers from other sources to your questions?

If so, do you mind sharing those answers?

Thank you.
 
None the wiser

Sorry dynamicnet, I am none the wiser yet.
I would think that this question is one that would come up frequently to SWsoft and I am hoping that one of the moderators of this forum will post a reply.
 
We do not yet have scalability docs ready.
However, the ballpark estimations for average usage is 5000 of users per modern server on Windows and 20000 users per same server on Linux.
Actually, the picture at usages higher than 5000/20000 users is much more complicated since you would need a cluster consisting of different machines:
- web front-end machines, optimized for HTTP serving
- file servers, optimized for fast & large HDD access with low RAM consumtion
- SQL servers, having small but fast HDDs and RAM enough to process the SQL load
- and of course the servers to keep the user sites load.

I would recommend you contacting sales representative in case of a large-scale deployment, so we could provide a sensible hardware estimations.

P.S. Recommended strategy for large projected loads is starting with a mini-cluster (three boxes, could be even VE's for a less expensive launch) and monitoring actual resources utilisation for upgrade decisions. Noone can predict actual behavior of your users with a reasonable level of confidence, so you'll either end up underestimating the hardware or overestimating it.
 
Back
Top