There is a known SharePoint bug which can affect some Nintex Workflows and cause them to freeze under certain conditions.
How the bug is defined
We believe the root cause of this issue is a fault with the SharePoint workflow engine. The bug causes workflows that have delay activities to fail and then wake up from a paused state and lose their workflow context.
When a parallel block is in another parallel block and one branch sets a variable and the other reads it, the reading branch doesn’t read the latest value of the variable.
Having a delay inside a parallel (in your example this is a task reminder) OR “while” loop (which are both Windows Workflow Foundation activities) suffer from the issue.
We believe this issue is caused by a race condition and is therefore, intermittent. However, this is exaggerated when the servers that are running workflow are under load. The issue is that if a workflow process is hydrated in the w3wp.exe process (i.e. web front end) and simultaneously hydrated in the owstimer.exe service, and both persist at the same time, instance data from one of the hydrated instances is lost. This leaves the workflow in a state where the user thinks the workflow is still ‘In Progress’ and certain workflow activities fail to complete / or run successfully.
We believe the fault is caused by the behavior of
- The SQL procedure proc_UpdateListItemWorkflowInstanceData and;
- The calling method Microsoft.SharePoint.Workflow.SPWorkflowManager.SaveInstanceData.
Nintex Workflow actions that may be affected
Lack of consistency
This bug is intermittent, and for the most part, unpredictable. It is not always possible to tell when it will occur or will not and simply using one of the above actions by itself doesn’t necessarily mean you’ll encounter the bug.
Troubleshooting and Bug Fix
1) Download the Microsoft SharePoint August 2011 CU
If you believe this is the issue you are encountering, Microsoft has released a fix for the SharePoint workflow bug.
The August CU is available for Microsoft SharePoint Foundation 2010 (MSF) here:http://support.microsoft.com/default.aspx?scid=kb;EN-US;2553050
The full server package for SharePoint Server 2010 is available here: http://support.microsoft.com/default.aspx?scid=kb;EN-US;2553048
It also contains the MSF2010 fixes so you will need only this one package if you are running SharePoint Server 2010.
The CU contains a number of bundled hotfixes, and the one that applies to this error is 2553031 http://support.microsoft.com/kb/2553031
2) Download the Microsoft SharePoint October 2011 CU
3) SharePoint timer service issue
If you are still experiencing issues, it may not be a result of the SharePoint bug but an issue with the SharePoint timer service. Please see this article Workflows not continuing - Timer Service and Delays
We are experiencing this issue in SP 2013 as well. Is there a fix for this in 2013? A business can't run on Nintex if it is prone to these kinds of unexplained failures!