Job Statuses
Across its lifetime, a job instance has various statuses, as shown in the Status column in the Running Jobs table below. To find out more about job statuses, see the subsection below.
Job instance stages
The diagram below illustrates how a job instance changes from one state to another across its lifetime. It is assumed that no loss of FlowForce Server service or network interruptions have occurred. Note that some of the statuses take a very short time span and will not normally be visible in the user interface.
Job statuses can be divided broadly into two types: created and finished. Each of these types is further divided into different statuses (see below).
Created
The Created status is the first state the job is in before any other action takes place. This status is abstract (i.e. it cannot be entered) and cannot be observed. The Created status is a superset of the following statuses: Starting, Waiting, Waiting for slot, Running, and Aborting (see details below).
If the execution queue has an opening and the instance is not delayed for some reason, it proceeds to the Starting status. The Starting status has a short time frame and lasts while the instance starts up. Then the job instance usually switches to the Running status.
|
If the instance is delayed, it receives the Waiting status.
|
If the job instance is ready to run, but the execution queue is currently full, this job instance switches to the Waiting for slot status. An execution queue has a limited number of slots. Therefore, only the specified number of job instances can be executed in parallel in the same queue. For details, see Queue Settings. Any further instances arriving for that queue will wait until a slot becomes available.
|
Indicates that the job instance is currently running and will stay in this state until the execution is complete or until some external event occurs that ends the execution prematurely. Except for a very brief time window at the beginning, this status has a step number associated with it. Therefore, the instance gets the Running step {step} status. Job instances can also have the following statuses: Running postponed steps and Running postponed step {step}. To find out more about postponed steps, see Postponed Steps.
|
A job instance switches to this status when the user cancels a job. It might take FlowForce Server some time to process the request. The Aborting status acknowledges the receipt of this request. Note that the job instance may actually be able to complete successfully before it switches to the Aborted state. If this happens, the job will be reported as having finished successfully. If the previous status had a step number, the Aborting after step {step} status would be shown instead of Aborting.
|
Finished
The Finished status is abstract (i.e. it cannot be entered) and includes the following statuses: Finished successfully, Failed, Aborted, Interrupted, Superseded, Lost connection, Synchronizing, Untracked, Recovering (see details below).
This is a final state which indicates that the job has finished successfully. The status Finished successfully after step {step} additionally indicates that the successful completion is associated with a particular step number.
|
The execution of the job instance has failed. This is a final status and there will be no further attempts to run the job instance. The Failed after step {step} status additionally indicates that the failure is associated with a step number.
|
This status indicates that a user has stopped the job although it may also happen indirectly after an unexpected shutdown. This is a final state that indicates that at least some part of the job has not finished. If the previous status had a step number, the Aborted after step {step} status would be shown instead of Aborted.
|
The execution of the job instance has been interrupted. This is a more forceful variation of the Aborted state. The job instance cannot be restarted. Therefore, it should be treated as failed. To avoid data inconsistency, it is recommended to check the outcome manually.
|
This status means that the job instance has not executed anything and that some other instance might have run instead of it. This status can only appear before the Starting status when, for example, the triggerfile has changed again during the settle time specified by the Wait N seconds for settle option. The Superseded status is not a critical condition.
|
This status applies when multiple FlowForce instances run as a cluster. This status indicates that the master machine has lost connection to the worker machine. When the connection is lost, FlowForce Server does not know whether the instance is still running. When the worker connection is reestablished, the instance switches to the Synchronizing status.
|
This status applies when multiple FlowForce instances run as a cluster. In a clustered setup, the master machine gets the current progress of job instances from the worker machines. When the worker connection is reestablished, the instance starts synchronizing and FlowForce is trying to get the latest status from the worker.
|
An instance gets the Untracked status when FlowForce crashes or is terminated while the instance is still running. You have to abort such a job manually by clicking the Stop job button in the Running Jobs list; otherwise, it stays in that list until the next FlowForce service restarts. When you stop an untracked job, it goes to the Aborted state and remains in the list for some time (currently about 90 seconds).
|
When an instance has become untracked, FlowForce Server will switch on the Recovering state before the job instance can proceed.
|