Another way would be to change the test to wait the failover... or change the ServiceContainer properly to setup the timeout on the test.
Unless I'm mistaken, I think this short timeout is only valid on a test scenario, hence change the file that would be used on production is not suitable.
actually before last release (cp04), the effective value was:
- <attribute name="validatorPingPeriod" isParam="true">2000</attribute> - <attribute name="validatorPingTimeout" isParam="true">1000</attribute>
because those parameters have never been passed to remoting layer. So it should be fine.
But there are some compaint about this, as it will affect tests relying on the current timeout value, so we should consider changing them to reasonable values after this cp release.
2 seconds for the ping is too short.
That should only happen on tests... please don't deliver the release with that change.
I updated the configures (ping period and ping timeout) to be equivalent to previous values. So test relying on related timing will not be affected by those values.