2 people found this helpful
This could also be due to the theme not being synchronized to all nodes so that users are hitting a node without the new theme. Can you open a support case for this? I checked out some cases and in the past the issue was the time between propagation of the theme to all nodes not having a long enough window. The solution was to alter the System Tasks page for Configuration in the admin console.
Here is a screen shot:
And the note from engineering was: "Looks like I got the theme to propagate across all 4 nodes. I ended up having to increase that setting to 120000 ms (2 minutes); however I've backed it back down to 60000 for now."
So I would err on the side of caution, open a case, and see what support says.
Thanks a lot for that response. It makes sense. I had opened a support case for this and updated it based on your feedback. Ill update when i hear back from them.
1 person found this helpful
And a little more explanation:
Everything seems to be pointing to the size of the zip file and the time it takes to process/transfer it between each of the webapps and the database and staying within the process window for the cluster messages that we've seen.