File Monitoring

[Joey]

I've had some issues of file monitoring periodically halting for Data Pump. I have been restarting the Data Pump Server awhen I notice input files that should have been processed, but have not. The processes run with no problems after restarting the server. I receive the error message listed under Monarch Data Pump Events every time. Has anyone encountered this message before? Any ideas on what might be causing the file monitoring mechanism to halt?
Thanks.

DwchServer.InvalidFindNextChangeNotificationExcept ion: Error setting up a next change notification for **Monitored Location*** - error 64
at DwchServer.ChangeNotification.Next()
at DwchServer.e.a(Int32 A_0)
at DwchServer.e.d()

[Gareth Horton]

Joey,

I think the problem here is that there has been a loss of connection to these network resources, either the network has dropped (even for a second) or perhaps the server on which the folders are located has been rebooted etc, which has killed the thread monitoring the locations.

Note that the monitoring will only be broken for the location that lost the connection (obviously if the network goes down, that will be all of them, but if a server has been rebooted, then folders on another server will not be affected)

As you rightly state, restarting the service restores the monitor threads.

We will be implementing a mechanism in the next version that means a monitor thread will report back if it loses its connection, and we will retry that resource until it comes back on line.


Gareth

Originally posted by Joey:
I've had some issues of file monitoring periodically halting for Data Pump. I have been restarting the Data Pump Server awhen I notice input files that should have been processed, but have not. The processes run with no problems after restarting the server. I receive the error message listed under Monarch Data Pump Events every time. Has anyone encountered this message before? Any ideas on what might be causing the file monitoring mechanism to halt?
Thanks.

DwchServer.InvalidFindNextChangeNotificationExcept ion: Error setting up a next change notification for **Monitored Location*** - error 64
at DwchServer.ChangeNotification.Next()
at DwchServer.e.a(Int32 A_0)
at DwchServer.e.d()

[Joey]

Thanks for your response. We recently migrated Datapump so it was on its own server. Unfortunately the new server goes through a router instead of a direct connection to the fileserver. The router is recycled every Sunday at midnight. Brief interuption in the network, but causes Data Pump to crash. We've been spinning our wheels for months before your suggestion. Again, thanks!

We look forward to the next version to have a fix for this!

[Joey]

We've also recieved an an error:

DwchServer.InvalidFindNextChangeNotificationExcept ion: Error setting up a next change notification. Error 56. This is accompanied with:

DwchServer.CannotLoadXprjException: Cannot load xprj file. ---> System.IO.IOException: The handle is invalid.

Note that I removed the file paths.

Similarly, restarting the server causes these to work just fine. Do you think this is also caused by a loss of network resources, or could something else be the cause of this?

[Gareth Horton]

Joey,

This is 99% related to the network outage issue.

8.5 is now available free to maintenance customers, so you should be sent a CD in the next few days if you are on maintenance.

In the meantime, if it is safe to do so (i.e you don't expect any files to come in at that time) you could create a batch file to stop and start
the Data Pump service after the router reboot on Sunday and run it via a scheduled task on the MDP server. This should refresh the monitor threads.

For example -

net stop "Monarch Data Pump 8 Service"
net start "Monarch Data Pump 8 Service"

Gareth
Originally posted by Joey:
We've also recieved an an error:

DwchServer.InvalidFindNextChangeNotificationExcept ion: Error setting up a next change notification. Error 56. This is accompanied with:

DwchServer.CannotLoadXprjException: Cannot load xprj file. ---> System.IO.IOException: The handle is invalid.

Note that I removed the file paths.

Similarly, restarting the server causes these to work just fine. Do you think this is also caused by a loss of network resources, or could something else be the cause of this?

[Joel F]

I'm having this problem also. This weekend file monitoring broke and we also have the 8.5 version installed. It would be VERY helpful if the errors Datapump put into the event log set the EventID to something specific to the error instead of being zero for everything. If this error 56 was set in the EventID we have a product called What's Up Gold that could monitor for this error and take actions to re-enable file monitoring.

Another strange thing is that I couldn't find the error 56 in the event log with the 8.5 version. Should I have still seen the error 56 with the new version?