Expand my Community achievements bar.

Dive into Adobe Summit 2024! Explore curated list of AEM sessions & labs, register, connect with experts, ask questions, engage, and share insights. Don't miss the excitement.

Unable to author Experience Fragment. Throws 504 gateway error

Avatar

Level 2

Hi,

I am on AEM 6.5 SP 7.  Whenever I am trying to author an experience fragment, I am getting below error. Using "Admin" account with full privileges.  Could not find anything in logs.. Has anyone faced this before? The content is getting updated when I refresh the page. However, the dialog is not closing on it's own and throwing timeout error with message "Fail to submit the form"

beast08_0-1648657631586.png

 

7 Replies

Avatar

Community Advisor

Hi @beast08 ,

 

Please check which request is giving a timeout. You can use OSGI console (https//<IP>:<PORT>/system/console/requests) to see the recent requests. I believe from your Experience Fragment you must be calling a servlet/service which is giving a timeout.

Avatar

Level 2

Thanks for the response.

 

It is timing out on SlingPostServlet - ModifyOperation: 

beast08_0-1648660030845.png

 

Avatar

Employee Advisor

@beast08 A bit unusual issue, can you try a similar thing In another server with the same permissions? I think it's something with the instance.

Avatar

Level 2

it's working in other server. This is instance specific. 

Not sure what's wrong. User permissions and other configurations are same.

Avatar

Level 2

We were able to fix the issue by re-indexing failing oak indexes.

 

It was really weird that failing oak indexes were affecting experience fragment authoring. 

Avatar

Employee Advisor

@beast08  that would have been my next suggestion, I expect a corrupted index to not stop but impact a fully working functionality. Glad you got it working!

Avatar

Level 1

We are facing exact same issue with our experience fragments on AEM 6.5 SP18 On Premise. 

Tried bunch of things still no resolution: 

  • Restarts - Instance and also few related bundles
  • Consistency check with came valid for all and no issues.
  • Reindexing entire repo
  • Investigated thread dumps - no sign of issues.

Any suggestions what else can we try to fix this problem ?