2

Running enterprise SDE on Sql Server 2008, ArcGIS 10.1, SP1. The issue is old locks on all feature classes within a specific feature dataset that aren't being released. When viewed with an admin connection through ArcCatalog, these locks were acquired over a year ago, and have no owner. View User comes up with nothing, and Disconnect User returns an error.

Any ideas on what could be causing these locks/why they are persisting in the seeming absence of a connection, and how to get rid of them?

Dan Jurgella
  • 2,378
  • 13
  • 15
  • related http://gis.stackexchange.com/questions/18958/arcsde-is-there-an-automated-way-of-releasing-all-layers-locks – Mapperz Jan 16 '15 at 21:25

2 Answers2

2

I'll start by saying this issue is a great candidate for a support call if that's available to you. However, you can get a better idea of what's happening on your own with the following:

Orphaned locks can be caused by application crashes or critical errors during some geodatabase operations. There is also an ArcGIS server bug in 10.1 when applying simultaneous edits via feature service causing these types of geodatabase issues.

http://support.esri.com/en/bugs/nimbus/TklNMDg1NTUz

When was the last time the Sql Server service was restarted? Has it been longer than the date seen in the Administer Geodatabase dialog? If so, try bouncing the service and check the locks again.

If you can confirm that there are no users currently making valid connections you can query the SDE locks and process_information tables to get a fairly accurate view of potential orphaned connections and locks.

SELECT * FROM sde.sde_process_information;


SELECT * FROM sde.table_locks;
SELECT * FROM sde.layer_locks;
SELECT * FROM sde.object_locks;
SELECT * FROM sde.state_locks;
Ken
  • 269
  • 1
  • 6
  • 1
    Yeah, ESRI support is an option for us, and I'll go that route if I can't figure this out. The feature dataset is rarely edited, so I think the bug you referenced wouldn't be the problem. In querying the tables as you suggest I found the locks in the table_locks table. And in the process_information table was no corresponding connection. I'm not sure when the last time the service was restarted. I'll have to look into that and give that a shot. Thanks for the suggestions! – Dan Jurgella Jan 16 '15 at 21:00
2

We ended up having our dba go in and delete these orphaned locks from the table_locks table through the SQL backend. No further issues so far.

Dan Jurgella
  • 2,378
  • 13
  • 15