we installed the Portal 4 times on exactly the same environments (VMs), same files (Portal 22.214.171.124), same settings but 2 of them show no data in the Analytics.
We have no error and we see nothing unsual in the logs. I have seen cases where the Analytics do not work but it gives error when you load the Analytics page. In our case we have no errors. And looking into the logs, it actually seems that analytics data do reach the portal from the gateway. We see the below behaviour:
The Portal is up and running with an enrolled Gateway for about a month now.
Has anyone had similar issue?
We can see, by running docker service logs -f portal_analytics-server, the below for example:
Upserting 25 Jarvis Records to api_metrics for tenants [53b42703-170f-4b1e-9796-43b345cb9d0f]
Successfully upserted 25 Jarvis Records to api_metrics
Processed write thread: WriteThread-HMP7 with 25 records
Most of the time I have seen this is because customers are using Gateway published API's and there are some policies and updates that have to be made to enable analytics. These changes are documented in the docops production documentation. For Portal published this should not be necessary and the only thing that I can think of that would prevent this is communication problems between the gateway and portal that prevents the gateway runtime metrics from getting pushed to the portal.
thank you for your reply.
We have made the nessesary changes to the Gateway published APIs (as per the recommended doc). As menioned above we have 2 environments that work and 2 that do not. They have exactlty the same settings, policies etc. From the logs (example above) we can see that we have records going into Jarvis from the Gateway. So we suspect there is something else that doesn't happen between the docker containers.
Can you please check that there are no Jarvis containers restarting in those environments not working. Is there a substantial difference in the amount of data being pushed to the portal in different environments? Also, if you could include a Case # from Support I will help them where I can and we can provide solution here when fixed. Thanks!
I can confirm that we have no contaires restarting and the data is more or less the same.
There is a Case open #01320114.
All things being equal I suspect something is different with the updates to policies and/or perhaps process involved to do the update in these non-working environments. Thanks for the Case #. Cheers!