Your achievements

Level 1

0% to

Level 2

Tip /
Sign in

Sign in to Community

to gain points, level up, and earn exciting badges like the new
Bedrock Mission!

Learn more

View all

Sign in to view all badges

SOLVED

Dispatcher response taking more than 10 minutes for some clientlibs JS files

luismoren
Level 2
Level 2

Hello All,

Early this week we received reports of users that have challenges to modify Sites pages and CF in Author servers of lower environments. When people click on edit a page on any author site, keeps loading and loading up to 10 minutes and then the page is available for editing.
Looking into this error we start by identifying the files causing this delay and one of these files was metadataeditor.js under the path:

etc.clientlibs/clientlibs/granite/coralui2/optional/imageeditor.js

our first step was to bypass dispatcher and access directly to the author server using its IP. The result was the page loading correctly in a few seconds. We thought then that the dispatcher was the error and went to check latest changes that may be causing this, but no changes have been done in any configuration file in the last months under the dispatcher servers and is serving other files correctly, only some files specifically "some" under etc.clientlibs have this delay. we got back to the Author server thinking that maybe the clientlibs needed to be rebuild or cache removed but that didn't fix the incident. User access is not an issue since we replicated the same with admin account.

Going back o dispatcher we look into the logs and found that request is being received and allowed, then contacted the render(author) and retrieve the JS  normally as other files do, but somewhere between sending the file back to the client is where the delay happens.

 

I was wondering if anyone on the community have faced something related, our servers are under AMS and our Success Engineer is looking into a possible cause on a higher layer above dispatcher as the application gateway but any help is appreciated in case there are any suggestions of anything we can check or validate in the author or dispatcher servers to find the root cause in case is under this layer.

Thank you in advance for reading this question and any support.

1 Accepted Solution
luismoren
Correct answer by
Level 2
Level 2

Hello All,

 

Found the root cause and was not related to the AEM implementation but the VPN used to access to AEM.Since any access to our network require VPN login, we never noticed that accessing from outside the VPN actually loads all resources without any delay.

Thank you for your help on this thread, closing with cause "Not AEM related"

View solution in original post

3 Replies
bangar50
Level 7
Level 7

Hi @luismoren ,

          Please use clientlibs versioning   it will giv faster performance than normal.

Kr,

Sanjay

 

luismoren
Level 2
Level 2

Thank you for your response @bangar50 , is true that versioning can add performance and even solve known issues when JS or CSS files cached are not automatically flushed on web client after a code release, however this is not the cause of the error since the author was loading correctly and just start failinf from one day  to another and the author or dispatcher have not been found so far the cause of the incident, even the performance on them.

luismoren
Correct answer by
Level 2
Level 2

Hello All,

 

Found the root cause and was not related to the AEM implementation but the VPN used to access to AEM.Since any access to our network require VPN login, we never noticed that accessing from outside the VPN actually loads all resources without any delay.

Thank you for your help on this thread, closing with cause "Not AEM related"

View solution in original post