I am completing a Component Pack install of 7 from scratch as a POC install so single server and I have noticed the commands for setting Metrics and Type Ahead search has changed.
Not only does it now say version7 in the command but the port has changed:
SearchService.setES7QuickResultsBaseUrl("https://example.org:30098")
Previous:SearchService.setESQuickResultsBaseUrl("https://example.org:30099")
Is this correct as in the port? The reason I ask is that if it is correct I can't get it to work.
Both Metrics and type-ahead are not working and looking at the logs it is reporting that it can't connect over SSL on that port.
Any help would be great because I actually can't even find out how I look at what port it is listening on from my environment.
OK so I have got a bit further with this. Mainly this is my fault :) I used a set of previous 6.5 scripts I had created which had the old commands in it which referenced 30099 also I had not created the QuickResults index using the new commands for 7 either. It also tells you what the port is when you run the python config_blue_metrics.py command so I know it is indeed now 30098. Anyway I have done that now and Type Ahead is working, however Metrics is not working still and looking at the logs I am seeing this:
Elasticsearch W com.ibm.connections.metrics.elasticsearch.ElasticsearchClient httpRequest HTTP request to ElasticSearch failed. org.apache.http.conn.HttpHostConnectException: Connect to comp.acme.com:30099 [comp.acme.com/192.168.100.42] failed: Connection refused: connect
So it is still referencing the incorrect port. I am not sure how to change this. I can see the SSL Objects are correct in WebSphere. I ran the enableSslForMetrics command with correct port but still Metrics is trying to connect on the incorrect port.
Have I missed something or does anyone know how to change this so it is looking at the correct port for Metrics?
Hi Richard,
Since you initially deployed ElasticSearch using scripts that reference port 30099, I think you could change the port by upgrading ElasticSearch using the command documented here, only referencing port 30098 instead of 30099:
https://help.hcltechsw.com/connections/v7/admin/install/es_upgrade_helm_chart.html
Tony Dezanet
HCL Connections Support
Thanks Tony,
I ran that but its strange it loads the Metrics for People and Participation however the numbers are all set to 0. If I change the group or view details I get an unknown error has occurred. This is the same if I change to any other type of Metrics views.
If I then look at the logs I am seeing errors about the indexes not being there -
[22/03/21 13:15:33:412 GMT] 000001c3 Elasticsearch W com.ibm.connections.metrics.elasticsearch.ElasticsearchClient httpRequest ElasticSearch returned - 404 : {"error":{"root_cause":[{"type":"index_not_found_exception","reason":"no such index [null]","resource.type":"index_or_alias","resource.id":"null","index_uuid":"_na_","index":"null"}],"type":"index_not_found_exception","reason":"no such index [null]","resource.type":"index_or_alias","resource.id":"null","index_uuid":"_na_","index":"null"},"status":404}
I made the change you said a few days ago so I don't think this is a wait and see if the indexes get created seems like there is more of an issue here.
Hi Richard,
I think it would be best to open a case with HCL Support for a more in-depth analysis.
Tony
Thanks Tony, I have raised a call for this with HCL so see what they come up with.