5:35 AM
Hello world. I want to understand what is happening under the hood in my case. And, sorry, at the moment, I don't have a lot of details. All I know is the general architecture of the application. Our app talks to an oracle db behind a firewall.
The firewall has a policy to close idle tcp sessions.
When that happens, the application connection pool complains of a connection pool request timeout errors. In other words, the db layer in the application pools the request in its internal queue; if it sits there waiting for a connection (in the connection pool) for more than a minute, it errors out...
We observe this happening in two scenarios. (1) after a network disruption (between the application and db) is resolved; i.e. the link between the two re-establishes..., and, (2) if the application has been running for a long time
In both scenarios, the deployment team's solution was to restart the application. After which the application began to magically talk to the db all fine. So my question is...why did the restarting of the application actually help?
