Quantcast
Channel: SCN : All Content - All Communities
Viewing all articles
Browse latest Browse all 3523

Agentry "keep alive" setting

$
0
0

Hi everyone,

 

We are deploying SAP Work Mgr. 6.2 on SMP SP07. We have had issues with initial syncs when connecting via the Cisco load balancer in front of the 2 SMP nodes, syncs initiated by connecting directly to either SMP nodes were always successful.

 

After trouble shooting this with the load balancer experts, we found out that the load balancer was killing the TCP connection after 2 minutes. Due to the volume of data, the time it takes to update larger complex tables like equipment, functional locations, characteristics etc. will always be > 5 minutes. During this time, there is no communication b/w the Agentry server and the client as the server is waiting for data, this would cause the load balancer to kill the connection.

 

We increased the TCP timeout on load balancer to 1 hour and the syncs started going through just fine. We are now investigating the possibility of configuring "keep alive" in the Agentry application, currently it is set to 3600 and we are looking to reduce it to 90 secs.

 

keep_alive.JPG

 

 

I have the following questions:

 

- Am I looking at the correct parameter here? Per the description in the above image(from the SAP Help documentation), it seems like reducing it to 90 secs would force the server to keep the connection alive every 90 secs.

 

- Could there be any adverse impact from a performance standpoint if we reduce this setting to 90 secs?

 

- Is this setting part of any configuration file or is the only way to change it is via the management console under "App Specific Settings"?

 

Any help you can provide would be much appreciated!

 

Cheers,

Abhinav


Viewing all articles
Browse latest Browse all 3523

Trending Articles