Workflow error - padding and instance store

Permanently deleted user

03 Augusti 2016 15:03

We have recently moved our Sites and Workflows folders to a Shared Area Network. This works as expected however we are now receiving this error intermittently:

 

'Unable to load specified workflow from instance store.

 

Unable to load workflow running instance data.

 

Padding is invalid and cannot be removed.'

 

Does anyone have any experience of this?

0

• 

  Permanently deleted user

  25 Augusti 2016 20:34

  Hi Lloyd,

   

  Interesting!  I've never seen that before, and I can't find a record for it in our bug tracker.  I would encourage you to open up a support case so that one of our analysts can help you investigate this issue.  

   

  Thanks,

   

  Danny

  0
• 

  Permanently deleted user

  06 September 2016 13:22

  We've recently started seeing this error occur in our multiserver/load balanced environments after updating to version 4.5 in the last few weeks. Did you ever get this issue resolved?

  0
• 

  Permanently deleted user

  24 Oktober 2016 19:14

  We're seeing something similar... There was a workflow error running activity: Exception has been

   

  thrown by the target of an invocation.

   

  Unable to load specified workflow from instance store.

   

  Unable to load workflow running instance data.

   

  JSON string is invalid or not supported.

   

  Invalid object passed in, member name expected.

   

  Anyone ever get to the bottom of this?

  0
• 

  Permanently deleted user

  15 November 2016 16:48

  We recently encountered something similiar in running our workflows, after upgrading from Essentials Manager 4.4 to 4.6.1. This issue only occurs in our loadbalanced environment and not all the time. In our single server environment everything runs fine. Our quick solution was to switch off one of the essentials servers, that seems to fix the problem for now.

   

  It seems like the seperate workflow activities within one workflow are run against different essential managers due to the loadbalancer. This results in the "Unable to load specified workflow from instance store." errros. 

   

  In our 4.4 load balanced configuration this was no problem, both essential servers were not aware of eachother. In the newer version there seem to be some major changes in the way Geocortex should be clustered / loadbalanced. 

  0
• 

  Permanently deleted user

  22 November 2016 14:20

  Yes we have found a work around to be just run on one server. We are currently on version 4.5. We beleive the fix for this version is to implementing 'sticky sessions' on the server which will persist a users connections to the relevant servers. In 4.6 you should be able to cluster the GE servers together as part of the post install, although you may need at least three servers to do this.

  0
• 

  Permanently deleted user

  13 Juli 2018 13:47

  I know this thread is a couple years old but we just ran into this error yesterday after upgrading to a new load balancer.

   

  The solution for us was to enable persistance / sticky bit on the session so that the load balancer always sends requests to the same server (we have 3 load-balanced Essentials servers).  Hopefully this helps others that encounter the same error message.

  0
• 

  Permanently deleted user

  20 Januari 2020 20:29

  Peter, same thing for us with the load balancer and enabling persistence.

  0

Du måste logga in om du vill lämna en kommentar.