Hello,

This is in relation to an ongoing issue where nodes fail on the front end. This seems to be a discrepancy between the published token and whether any xml is stored for use in the umbraco.config.

See previous issue: http://our.umbraco.org/forum/using/ui-questions/35164-All-nodes-reverting-to-'hash'-url-472-49

Our client has recreated this issue when creating a new node and sorting this while in its unpublished state. This caused all nodes to be absent from the umbraco xml and cause all macros (navigation etc...) to fail on the homepage.

In such instances the rogue nodes appear to be published but have an empty '#' as the url.

Our previous answer to this was to republish all nodes - right click publish on the actual node, right clicking on the 'Content' node and 'republish entire site' does not work in this situation.

Instead of doing this we ran the following script on each parent node that was having this issue:

SELECT * 
FROM 
umbracoNode
FULL OUTER JOIN 
cmsContentXml on umbracoNode.id = cmsContentXml.nodeid
WHERE
umbracoNode.parentId = {parentNodeId}

This higlighted only one node that had missing xml. Despite others appearing to be in the limbo, hashed url state.

Publishing this rouge node and refreshing the umbraco.config fixed all macros.

Does anyone have a solution or working process to avoid this issue?

Our current avoidance is to implement a content freeze when dealing with any DocTypes, Macros, etc... on live either with courier or manual. Our issue now is that the nodes are failing when a content editor makes a change, which we believe to be sorting an unpublished node in this situation.

Any help would be greatly received,

Adam