Nintex workflows failing on start - cancelled by system account

  • 23 January 2019
  • 6 replies
  • 65 views

Badge +3

Hi,

A while back I raised a question regarding the .NET security update that stopped workflows running. We've recently released some updates on our SharePoint 2010 farm and this security update was included automatically.

We were able to remove and restart the servers which seemed to fix the issue - all workflows appeared to be runnning again.

However, since then I've had some strange intermittent issues that I cannot rectify:


1. Initially, when trying to open a nintex workflow I recieved this error (this has since stopped happening)

2. The same workflow is now failing on start, every new item that is created does this regardless of the initiator. Strangely, this is only affecting this workflow - I have tested three others in seperate lists & sites that work fine.

The workflow only has 4 or 5 actions in it, it pauses for 1 minute, then updates a column based on whether another is empty or not. The list isnt huge either only a couple hundred items so im unsure what the cause is. Also worth noting the workflow worked fine prior to the earlier mentioned security updates as of last week!


Any help would be greatly appeciated.


Thanks


Anthony


6 replies

Badge +3

UPDATE: have since removed the .NET security update that had been auto installed on one of the WFE servers in this farm. Restarted all servers in the farm (app then wfe's) but workflows are still failing on start.

Badge +3

Have you deployed the fixes for the SharePoint Timer Service also?

Because of the 1 minute pause you have in the workflow, the workflow is going into the SharePoint Timer Service, and if the web.config for the SPTS has not been fixed, it would explain the second error.

If you have multiple servers in the farm, make sure to deploy the fix to all servers.

https://blogs.msdn.microsoft.com/rodneyviana/2018/09/13/after-installing-net-security-patches-to-address-cve-2018-8421-s… 

Badge +3

Hi Arne,

We havent deployed any fixes in our environment, the last time this happened in October we couldnt get those fixes to work so we removed the security update and rolled the patches back, restarted and workflows have been working ever since.

This is what we were looking to do this time, but so far it hasnt worked.

Thanks


Anthony

Badge +3

Then I assume some of the damage done by the patches is still making life difficult for you. Take a look at your web.config files, with reference to the article linked above.

Badge +3

I think we are thinking along the same lines happy.png I've just made this suggestion to the infrastruture guys myself. I can only think the web.config must be missing those dependancies in the article.

I'l report back our findings happy.png

Thanks

Anthony

Badge +3

A final update to this saga from me hopefully, we ran the Add-CodeDomAuthorizedTypeToOWSTimerConfig.ps1 script as recommended to fix the OWS timer which worked! All our workflows are now up and running again. As noted, we did notice some odd behaviour regarding nintex workflows with pauses either failing on start, or erroring but completing before we ran this script. But, the script above fixes this in any case

 

Thanks

Anthony

Reply