ShareAspace host
Stopping the ShareAspace host
To shutdown ShareAspace, stop the Microsoft Windows Service ShareAspace Nova Host.
PowerShell
Stop-Service -Name "Eurostep.SAS.ServiceHost"
Shutdown process
If there are running DataExchange jobs when the ShareAspace host shutdown is initiated:
- Jobs that are running but are not in a commit stage will be canceled.
- The shutdown will wait for any running commits to finish before the host is stopped.
- Jobs that are on a queue will remain in the queued status. Processing of these jobs will continue once the host is started.
Tip
For DataExchange jobs that are dependent on the order in which they are run it is recommended to stop all DataExchange queues and let the running jobs finish before shutting down the system.
Tip
Before shutting down the ShareAspace host, it might be preferable to either stop the Microsoft Internet Information Services hosting the ShareAspace web application or to redirect users trying to access to web application to a "system is currently down"-page.
Starting the ShareAspace host
To start the ShareAspace host, start the Microsoft Windows Service ShareAspace Nova Host.
PowerShell
Start-Service -Name "Eurostep.SAS.ServiceHost"
Startup process
At startup ShareAspace will run a series of consistency checks on the collection database, all space databases and all indexes for each space.
If the ShareAspace host was shutdown "gracefully" ShareAspace will be reported as started and any active DataExchange queues will continue to schedule queued jobs.
If the ShareAspace host was unexpectedly shutdown, due to a hard shutdown of the ShareAspace host process or due to a hard system failure, and there were DataExchange jobs running at the time of the unexpected shutdown.
- The jobs that were running will be marked as failed.
- Provided that the system can repair automatically in the consistency stage (*), the DataExchange jobs will continue processing the jobs in the queue.
Note
(*) During start-up, the system does a full consistency check. First it verifies that all recorded transactions are available in the transaction journal. It also verifies all the checksums for the journal. The store then enters a recovery state where all transactions that had not been applied into the data file during normal operation are applied from the journal to the data file. This will guarantee that all journal data is processed and that all committed transactions are available after start-up. If any of these steps fail, the system will report this as a critical error and will terminate the start-up process.
Tip
For DataExchange jobs that are dependent on the order in which they are run it is recommended to have the queues configured with the reactive stop flag. This flag would, in this scenario, force the queues to stop automatically once the startup process marks the jobs as failed. This to allow an administrator to reschedule and reprioritize the failed jobs within the affected queues before activating them again.
Note
The startup time after a system failure will be much longer than the normal startup time. This is due to the consistency and repair stage.
Note
Depending on the server failure that caused an unexpected shutdown, ShareAspace might not be able to repair its state. In this scenario one need to restore the system, either by restoring a checkpoint or by doing a full system restore (snapshot install and index rebuild).