HTTP ERROR 503 ServletResolver service missing, cannot service requests

[kevingtan]

Hi All,

I am having a 503 error message after copying the content under /crx/de, say, cloning a template and modifying the sling:resourceType and jcr:title for the new template. This looks like a repetitive and consistent error for my dev site if I do a similar cloning, which ends up that I would have to wipe out the whole repository and rebuild it again. I am quite concerned about it because this kind of recovery can't be reproduced in an actual production environment or it will be very costly. I followed all kinds of successful remediation suggested in the community, but haven't had any luck at this point.

What I have done so far for it is:

deleting all indexes, all kinds of xxx.lock files, /repository/launchpad/..../repoint.

I also looked into this post, https://cqdump.joerghoh.de/2022/03/17/how-to-analyze-authentication-support-missing/.

but it seems to be a general direction about the repoint. I see some other members having the same issues and without luck by following the same instructions as mine. Is there someone having a definite solution to this? 

Thanks!

-kt

[BrianKasingli]

Are you trying to CRUD, create read update delete, crx/de with the JAVA APIs? There might be some error in your code where it wipes out specific parts of AEM that you are not intended on, for example, wiping out the entire /apps, /etc, content. How are you proceeding with these copy-and-paste actions?

Secondly, you cannot copy and paste pages (while they are tightly coupled to a template), because your root page, allowedTemplates are not set up in the page properties. Ensuring that you have allowedTemplates setup will guarantee that you can create or copy pages from one folder path to another.

[kevingtan]

No, I just went to /crx/de, copying an existing template under /apps/mysite/templates, pasting within the same directory, then modify the name and jcr:title. No code was involved, actually every time is the same and almost get the error every time. 

I was expecting to see the cloned template while creating a new page. It didn't show up so I restarted my bundle, no luck. Then I ran the stop and start script, the error showed up. Restarting the bundle via /system/console won't create such an issue, only restarting the aem service via bin. 

[DPrakashRaj]

Offline compaction and data store garbage collection will help you in solving 503 missing authentication exception.

the thing that you explained above leads to conclusion that you cling operations is corrupting your repository so would suggest you to check your cloning activity about the negative results 

[kevingtan]

We did weekly compaction in 6.0, not anymore since 6.5 has an online mechanism for it. Maybe the offline oak run provides some advantages that the online one doesn't have. Last time I tried it and it took too long and gave it up. But it looks like something I can't skip.

[DPrakashRaj]

What do you see in error.log?

[kevingtan]

Just like the screenshot, ERROR 503...etc. 

[aanchal-sikka · Accepted solution]

@kevingtan 

Template development should not be done on server. It should rather be done via code. 

On local:

1- Please use the copy option from the AEM's template UI.

2- Bring it to local IDE.

3- Make the necessary changes and push the details.

Else, if everything is done on local AEM, and working fine, then create a package of template and deploy on server

[kevingtan]

Thanks for the suggestion. 

[kevingtan]

Unfortunately, this doesn't solve my problem. I was skeptical about the idea of not implementing the templates under /crx/de since the official Adobe tutorial will instruct that how to create template nodes under /crx/de. I created an article using the AEM dashboard instead of using /crx/de, then deleted it within the dashboard also. When I restarted the AEM service, it failed to start. 

This doesn't happen on classic mode though. The error from the log is mssing 

org.apache.sling.engine.impl.SlingRequestProcessorImpl.

Any idea for this servlet?

Thanks!

[Love_Sharma]

Just want to check if you have tried deleting repoinit folder from AEM instance. Either you can run below command

 rm -rf crx-quickstart/launchpad/config/org/apache/sling/jcr/repoinit OR delete the folder manually. 

[kevingtan]

Yes, I did. I deleted the repointit, index folder, and all the .lock files, after shutting down the server. 

[Vishal_Waghmare]

Hi @kevingtan, were you able to resolve this issue? Re-install AEM instance every day only option I am currently left with.

Please let me know if you found the solution

Thanks

[kevingtan]

It seems that there was an xml error in our config file. If there is spacing in front of users in the User Mapper Service config, it will break after adding a new service user. I had the error quite frequently, and I haven't had such an issue anymore after trimming the space. Or it can be luck.