The past few days while putting on the finishing touches of a workflow I've been developing for 3 weeks, I am getting this frequent error message "operationId is required".

In the Instance activities, the last row I'll see is either: Update items (Waiting) or Retrieve an item (Waiting) with a Status of Failed in red. Expand it and I get 'Detailed error message is unavailable for this action.' No other details are given.

This seems to be related to only SharePoint Online actions. It happens either during SharePoint 'Retrieve an Item' or SharePoint 'Update Items'. (I'm only updating 1 row though, the row the flow was triggered by). This is a hit or miss situation, sometimes the actions work fine and move right along to other parts of my flow. The errors happen at various different SharePoint actions within my flow design so it's not always the same action failing. I have about 10 SharePoint actions at different areas in the flow, all running sequentially so it's not an issue of Parallel branches trying to access SharePoint at the same moment. Occasionally the entire flow even finishes! Most times it's now failing. I have a couple of test SharePoint rows that the flow originates from, they all seem to fail at irregular moments. I've created other test data rows with very simple data in the columns to reduce the complexity, so it's not a data issue. Sometimes the failures even happen on a simple 1 column boolean field update.

I have full access to the row in SharePoint so it's not a permissions issue. The flow is initially triggered on a SharePoint list row insert or update. (IN fact, occasionally Nintex seems to miss the update in SharePoint so it doesn't even start a flow instance, which is extrememly disappointing because the reliability of this softare is now in doubt.) When it does run, the flow asks for data from users via forms and then updates the SharePoint row with that form data. Other forms are then routed to others and more data collected and SharePoint updated again. Always the same row in SharePoint.

I thought maybe the flow was running too fast and the SharePoint row was temporarily inaccessible, perhaps locked, so I tried adding a 'Pause for Duration' action of 5 minutes after every action to SharePoint, so I wouldn't hit SharePoint too frequently and give it time to commit any updates I'd previously sent. But, that is not helping. I am currently the only user working on rows within this SharePoint list, so it's not a multi-user concurrency issue. There are also no other flows running on the list row in SharePoint, I'm making sure that only one flow is active at a time.

I worked with Nintex on-premise in SharePoint 2013 for 8 years and it was a fantastic rock solid product. I loved it. So far, NWC has been disappointingly buggy and much more difficult to work with. I was just about to publish this flow and now it's unusable. Please help me get through this issue! Thank you so much.