Introducing: SequentialProcess

Ever tried to chain multiple calls to FdoUtil.exe using a batch file?

I did recently on a client site, and found out that for one reason or another the Task Scheduler feature of Windows Server 2008 decided that the batch file I wrote was not going to run. After trying to find out the reason why (with no success), I tried running the individual calls to FdoUtil.exe with the Task Scheduler. Lo and behold they ran!

So armed with this knowledge, I have introduced a 3rd task type into FDO Toolbox, which will be in the next beta release: The Sequential Process.

A sequential process is simply an XML document that contains a series of calls to FdoUtil.exe, with an option at each invocation to abort the process if the previous FdoUtil.exe invocation returns a non-zero result (FdoUtil.exe returns 0 on success, otherwise it returns one of the predefined values here)

So for example, you can define a sequential process like so:

1. Run a SQL command to delete some rows

2. Execute a bulk copy task file (eg. SDF to SQL Server)

3. Run a SQL command to rebuild indexes

The SequentialProcess document would look like this:

Here's what it looks like visually (yes! This will have design support):

When you run this sequential process. Each step will be run as a separate FdoUtil.exe process, with the root process waiting for a return result from this process before continuing onto the next step.

For reference, FdoUtil.exe currently supports the following commands:

• ApplySchema
• CreateDataStore
• Destroy (data store)
• DumpSchema
• CreateFile (data store)
• RegisterProvider
• UnregisterProvider
• BulkCopy
• RunTask
• ExecuteSql

The sequential process task has opened up the design space for more FdoUtil.exe commands. So this list will definitely grow in size in future releases.

And yes, the RunTask command now supports SequentialProcess documents, so theoretically speaking, you can also nest Sequential Processes and other supported task types.

As I said before this is task scheduler friendly (the main reason for introducing this feature), so you can hook this up via the task scheduler for painless daily data updates or other automated spatial data maintenance.