Troubleshooting SharePoint Event ID 6393
This event ID means “a SharePoint timer job had an unexpected failure” in short. Admittedly this isn’t overly informative given how many different timer-jobs there are for any given SharePoint farm. Your first challenge is therefore figuring out which job it’s talking about which shouldn’t be too hard given it gives you the class name at least.
Here’s some example class names just to give you an idea of what can cause this event ID to show up:
Let’s take the last one specifically though as it’s the most common in SharePoint 2013 – if you’re seeing another timer-job fail then this article probably isn’t for you. For some reason this timer-job doesn’t have the most thorough in-line exception handling so the timer-service catches it and reports to the event-log:
The Execute method of job definition Microsoft.Office.Server.UserProfiles.LMTRepopulationJob threw an exception
This timer-job handles the repopulation of feed cache for users, in short. If this fails you’ll probably not see any social feeds working. Here’s a small working example:
(That’s me and my cat when she were just a wee puss in that photo)
The fact you can see stuff in that list means the timer-job is working. A broken newsfeed might not even go noticed if you don’t use that part of SharePoint 2013 but that won’t stop the timer-job from running anyway, causing an error every time it runs regardless.
Fixing The Newsfeed & Event ID 6393 for LMTRepopulationJob
The newsfeed repopulation depend on a couple of key factors:
- A healthy AppFabric cache-cluster.
- The right permissions in the right places.
Quite often, at least one of those two dependencies aren’t right which will cause a flood of event ID 6393.
Ensure a Healthy AppFabric Cluster
Make sure AppFabric has no more than 3 instances running on the farm, preferably on the same machines as your web-front-end(s) as that’s where AppFabric is the most heavily needed, and make sure the cluster is healthy by following this super-awesome guide.
If you don’t see a nice & healthy cluster then just don’t continue
Check UPA Service-Account Permissions
There’s a couple of things you need to check here – does the user-profile service-application user have access to the cache cluster? Run this to check:
Grant-CacheAllowedClientAccount -Account "sfb-testnet\svc_user_profile"
Change “sfb-testnet\svc_user_profile” for the username running your user-profile application of course. Next, and this is a slightly odd one; make sure the user profile service-app user has the right permissions to the user profile service application itself.
That’s it; if I’ve missed anything I’ll add it later. You should now not see any more errors with event ID 6393.