public class StopSchedulingTaskOnClientOrder extends SchedulingOrder
When a client throws this from a TASK, it can still be scheduled with other tasks for which it is eligible.
This exception can also be thrown from a batched INITTASK or CLOSETASK. In this case, all eligible clients are scheduled repeatedly with the batched task until all clients executing it have thrown this exception. Then the next task in the list is scheduled onto eligible clients.
For example, suppose a test wants to run a fixed workload on each client, say N iterations of an operation. Each client throws this exception when it reaches N. To avoid doing all N iterations in a single task, it does the work in batches, and uses a HydraThreadLocal to keep track of how many it's done. That allows totalTaskTimeSec to be set arbitrarily large, but still keep maxResultWaitSec small to detect hangs early in long-running tests. If there are multiple workloads, be aware that a large totalTaskTimeSec means that termination must be achieved some other way, e.g., via additional stop scheduling orders or maxTimesToRun task attributes.
| Constructor and Description |
|---|
StopSchedulingTaskOnClientOrder() |
StopSchedulingTaskOnClientOrder(String s) |
getOrigin, getRootCause, setOriginaddSuppressed, fillInStackTrace, getCause, getLocalizedMessage, getMessage, getStackTrace, getSuppressed, initCause, printStackTrace, printStackTrace, printStackTrace, setStackTrace, toStringpublic StopSchedulingTaskOnClientOrder()
public StopSchedulingTaskOnClientOrder(String s)
Copyright © 2010-2015 Pivotal Software, Inc. All rights reserved.