Are there any operational best practices guides such as restart?

[younghunyun]

Hi

We use currently using Apigee, but we have used an other API GW product before. The other API GW product have known issues like memory leakage, there is an offical recommendation for operation to restart components of the product at least once per year.

Up to now, Apigee has satisfactory performance compared to previous API GW product and there are no operational problems. Our operation team wonders if there are any operational recommendations such as restarting components for manual operation tasks.

For example, if CPU OO% or Heap/Memory OO% or more occurs, restart the component of the node is recommended. Or recommend restarting routers, message processors, ... components at specific intervals like 6 months or 1 year.

The operations team, who had operated previous API GW product, is wondering about the cases for handling restarts as an emergency measure.

Of course, if there are no issues, it is good to use Apigee for our operation team without restarting for a long period of 3 years or more, but I am asking a question because I can't find the relevant information on the docs page.

[dchiesa1 · Accepted solution]

I think you are running Apigee Edge private cloud, AKA "OPDK". 

If so, start here in the documentation section entitled "After the Install". https://docs.apigee.com/private-cloud/v4.51.00/after-the-install

There is no fixed amount of time during which Apigee recommends restarting servers, as far as I know. There are specific recommendations on what to monitor.