Problem Number: 584

Affected System: OnlineDMP, DMS, DCS

Issue summary

When a client agrees a budget using the online portal it shows as agreed in DMS but there is no Budget information on DCS. This cause client dissatisfaction as there recent budget information is not available.  

Current Workaround

This error is picked up by the Review Teams Daily report’s which advisors investigate.

If a client calls apologies to them for the inconvenience and make them aware this is a known issue that is being investigated by the IT team (Application support).

Its Important that a ticket is raised regarding this via fresh.

Application support actions:

Run the attached script.

After testing with ROLLBACK in daily, take the following steps:

1) Change back to COMMIT and save the file in this location:   \\VM01DCSPRODBA01\DataFixes.  

2) After a short while, the file will be moved to \\VM01DCSPRODBA01\DataFixes\Processed. Open the file from this location - it will have some output added in a note at the bottom.

3) If the output shows the script failed, investigate why and resolve any issues. Contact the DBAs for further guidance if required. If the output shows a successful run, save this file again but in this location:  \\VMDCSPRO2DBA01\DataFixes

4) After a short while, the file will be moved to \\VMDCSPRO2DBA01\DataFixes\Processed. Open the file from this location - it will have some output added in a note at the bottom, beneath the initial output.

5)  If the output shows the script failed, investigate why and resolve any issues. Contact the DBAs for further guidance if required. If the output shows a successful run, copy both output notes from this script onto the ticket.

6) Confirm the ticket has been worked with the user and close the ticket.

Permanent Solution / Accepted Workaround

 The above work around the current process while this is under investigation. The latest work on this identifies that:

  • There are no DCS budget messages being logged to the DCS trace file. DMS has these.
  • The DCS serve is entering a zombie or unresponsive state. When a WCF service is in a "zombie" or unresponsive state, it typically means the underlying host process (like IIS or a self-hosted ServiceHost) is no longer processing requests, but the process itself is still running. Recovery generally requires restarting the service host process and implementing robust error handling to prevent the state from occurring again.

Solutions are being investigated.