If the Scheduler service has failed to start, there are two known scenarios that may be the cause.
There is either a dependency on SQL not starting prior to the service, or there is an issue with Disk Space on the Discovery Center server.
Below we discuss these in greater detail, and provide options for resolution.
Service fails to startup after system restart
If the host system is restarted, you can encounter a race between the SQL and Discovery Center Scheduler services. If an initial database connection cannot be established, the Scheduler service can not start.
If you have systems where this could be happening, this article outlines how to make the Scheduler service startup dependent upon the SQL service status.
Disk space prevents Calculate Fields completion
If there isn't sufficient Disk space on the application server, the Scheduler service may terminate unexpectedly during index processing. This is usually triggered during the "Calculate Fields" phase of processing an index, where the Scheduler does not have sufficient disk space to update the Lucene index.
There are three different options you can choose from to resolve this:
- Add more storage space to the drive that is full.
- Alternatively, you can clean up the contents of the Discovery Center SearchData folder by deleting the index.data folder.
- Do not attempt to remove just the files within the index.data folder.
- Only remove this folder when there are no other processes running in Discovery Center.
- Deleting this folder will cause the Workbench Search data to be removed until index re-classification is performed.
- Typical installation will place this folder at: "Program Files > Active Navigation > Discovery Center > SearchData > index.data".
- Do not attempt to remove just the files within the index.data folder.
- Lastly, you can update the search index folder to a location with more space (often this happens because the default C: location is space constrained). This is outlined in the article linked here.