Kibana dashboard not rendred

Question

I'm running a dockerized elastic cluster composed of 3 master and 3 data nodes on AWS instances, using rsyslog and logstash, i collect and store syslog events on elasticsearch index. till now everything was good until trying to build a vizualization dashboard on kibana. working on kibana i notice that the request take too much time (even on a small number of syslog message 'less than 5000 records'). The following error 'Loading chunk 20 failed.' is repeated a lot.

The dashboard shows data once, and throws errors many times.

and

syslog event are stored as the following schema :

No error logs on both side, kibana and elastic server.
I tried many suggestion found on the internet :

• index re-create.
• re-deploy kibana on new EC2 instance.
• versions upgrade.
• machine ressource scaling.
• re-configure kibana.

Note : Currently i'm deploying the following version :

• ES : 8.10.0
• Logstash : 8.10.0
• Kibana : 8.10.0-1

P.S : same issue was also with version 8.9.0.

i'm ready to provide any other further info could help to resolve this issue.

I appreciate any help from you. Thanks in advance.

Answer 1

Did you try Stack Management - Kibana Data Views and deleting the Data View which you see as Index in Kibana-Discover?

If you do not have Lenses, Histograms etc. tied to your index, you can try deleting and recreating it again on Data Views.

Answer 2

good news : Issue fixed, and cannot believe that the cause of all this pain was Nginx :/ even (basic nginx config), Secret was at: setting header "connection" to "upgrade" everything works nice.

proxy_set_header Connection 'upgrade';

unset header, issue return.
Hope that help others suffering same issue.