Recommend server resource for REST API

[#1]

Hi,

As I know it took few second per transaction, but I cannot have do load testing for large scale.

Could you please recommend server resource for REST API?

If the customer required 30k transaction per hour. 

Currently I recommended

• 2 Web Servers (Web & API) : 12 Cores, 32GB RAM. (Support 150 Concurrent using DocPop + 30k calling REST API per hour)
• 2 Application Servers : 12 Cores, 32GB RAM.
• 1 IdP server.

THANK YOU

[AdamShaneHyland]

Hi Kobpong.

Think of the ApiServer like you would think of the OnBase Application Server.  I would size it the same way.  There is more information about Application Server sizing in the Application Server MRG.

Best wishes.

[#3]

Hi Adam,

My resource scale is refer from HardwareResourceGuide_EP2 with it's large scale 400-1000 Concurrent.

The requirement is 30k times call API per hour from 10-20 external systems with 20 service accounts.   (AVG 500 calls/Second)

Unlucky that I can't find there information about Application Server sizing in the AppServer_EP2 (MRG).

Can you give me page number or link to the article?

Thank you.

[AdamShaneHyland]

Hi Kobpong.

My apologies, the information is in the Hardware Resource Guide.  There is a section under the Large - 400 to 1000 Concurrent Users which notes the recommended size of your Application Servers.  Keep in mind that you can scale out your solution with more load balanced Application Servers if needed.

Best wishes.

[#5]

Hi Adam,

Sure it must have load balance between the servers.

The question is my Web Server (Installed Web Server & API Server) can be support 30k times calling REST API? Or shall I split the API server out of the Web Server?

Thank you.