"Can't connect to the server (Empty reply from server)"

45 views
Skip to first unread message

Marco Valentim

unread,
Jul 28, 2022, 2:15:44 PMJul 28
to AtoM Users
Hi everyone,

I hope someone can help me with this problem:

Yesterday, I started installing AtoM 2.6.4 on my Ubuntu 22.04. I downloaded all the necessary programs (PHP7.2, MySQL, Nginx, Elasticsearch and even Memcached). Then, when I opened "http://localhost" for installation and reached the 3rd step, I clicked on "Save and Continue" and an error message appears, saying "Can't connect to the server (Empty reply from server)". I already tried removing and copying again AtoM files into /usr/share/nginx/atom, but it still shows me that message... What should I do?

Thanks for the attention,
Marco Valentim

Dan Gillean

unread,
Aug 4, 2022, 8:48:03 AM (10 days ago) Aug 4
to ICA-AtoM Users
Hi Marco, 

I confess I can't remember which step in the web installer is number 3  - I think it's the "Configure search" step? Either way, my first suggestion is to check whether or not the relevant service is running, and try restarting it. For example, for Elasticsearch, if you've followed our recommended installation instructions, you should be able to restart ES with: 
  • sudo systemctl restart elasticsearch
If it's the Configure Database step:
  • sudo systemctl restart mysql
You can also try reloading the webserver with: 
  • sudo systemctl reload nginx
You can try manually returning to an earlier step in the web installer and seeing if it will complete this time - for example:
  • http://<host>/index.php/sfInstallPlugin/loadData
Other ideas: 

Make sure the permissions are properly set: 
  • sudo chown -R www-data:www-data /usr/share/nginx/atom
There are also other common maintenance tasks you can try like clearing the application cache, restarting PHP-FPM, etc. These are listed on our troubleshooting page: 
Finally, it sounds different from the problem you are encountering but there is a known issue where the web installer sometimes times out before completing. Fortunately there is an easy workaround for this, described here: 

Dan Gillean, MAS, MLIS
AtoM Program Manager
Artefactual Systems, Inc.
604-527-2056
@accesstomemory
he / him


--
You received this message because you are subscribed to the Google Groups "AtoM Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ica-atom-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ica-atom-users/9f44ce36-1ed4-4460-a7ce-1636947b9483n%40googlegroups.com.

Marco Valentim

unread,
Aug 10, 2022, 4:56:28 AM (4 days ago) Aug 10
to AtoM Users
Hi Dan,

Thanks for your support. Unfortunately, I still can't pass through the 3rd step, which is indeed the "Configure Search" step.
I tried those first three commands you posted and nothing really changed there...
I also followed your suggestion (which is also referred in the last link you posted) and it shows me a common error in this forum but that I never got: "Sorry, something went wrong. The server returned a 500 Internal Server Error."
I already tried using the troubleshooting page, but I still can't get forward that step... Again, thanks for the attention.

Greetings,
Marco Valentim

Dan Gillean

unread,
Aug 11, 2022, 8:11:24 AM (3 days ago) Aug 11
to ICA-AtoM Users
Hi Marco, 

If the issue is the search index configuration, there are a few additional suggestions for checking and troubleshooting Elasticsearch on this page: 
Can you try running the search:status task and checking cluster health, and share the results with us?

Additionally, how much RAM does your AtoM server have? Many Elasticsearch issues are caused by a lack of memory.  As noted in our Requirements documentation, we recommend the following as target minimums for a production-ready environment:
  • Processor: 2 vCPUs @ 2.3GHz
  • Memory: 7GB
  • Disk space (processing): 50GB at a minimum for AtoM’s core stack plus more storage would be required for supporting any substantial number of digital objects.
I would suggest that for a production installation, 4GB RAM would be on the lower end, and 2GB would be the absolute lowest value for a very small installation. You may still experience issues with that little RAM, as Elasticsearch will need at least half of it. 

In some cases (such as this forum thread), we have heard of users having enough general memory on the server, but not enough available to Elasticsearch - particularly, the Java Virtual Machine (JVM)'s heap_size value. JVM is an Elasticsearch dependency, and the heap size is the amount of RAM allocated to the Java Virtual Machine of an Elasticsearch node, which is configured by setting -Xms and -Xmx values. I believe the default value at installation for the heap size is 1GB.  If your AtoM server has 4GB RAM or more but you're still having problems and none of the above suggestions have helped, then you might want to try increasing the heap size to 2GB. As a general rule, you should set -Xms and -Xmx to the SAME value, which should be 50% of your total available RAM. A higher heap size will give your node more memory for indexing and search operations. However, your node also requires memory for caching, so using 50% maintains a healthy balance between the two.

For some further information and suggestions on this, including how to increase the JVM heap size, see: 
Let us know what you find! 

Cheers, 

Dan Gillean, MAS, MLIS
AtoM Program Manager
Artefactual Systems, Inc.
604-527-2056
@accesstomemory
he / him

Reply all
Reply to author
Forward
0 new messages