Question about 'ResourcesOnNonCentralNodes' load balancing rule

[Bo_PT]

Last week, when I opened Qlik Sense hub site, I found the 'Monitoring apps' stream disappearred. The Qlik Sense environment is multi-node. I did some research and found an article (https://support.qlik.com/articles/000049033) which said the 'ResourcesOnNonCentralNodes' load balancing rule should be disabled. I did that and the issue was solved. However, this Qlik Sense enviornments were set up this way (with 'ResourcesOnNonCentralNodes' load balancing rule enabled) 2 years ago and it has been working well since then. Why in a sudden it started to have this issue last week? We did not do any system changes/upgrade in recent months.

[RafaelBarrios]

Hi @Bo_PT 
If the two standard balance rules remain as they come out of the box

Note that the rule on the right side has the monitoring stream ID, you could check if for some reason this ID has changed

Also, you should check that all your virtual proxies facing users are linked to all other engines.

i think, i your case, your virtual proxy used by users are not able to use central engine

Hope this helps.

Best

help users find answers! Don't forget to mark a solution that worked for you & to smash the like button!

[Bo_PT]

Hi, @RafaelBarrios 

Thanks for your reply. We have two nodes (central and rim) and 3 virtual proxies: one central, one rim, and the 3rd one for external users. Currently the central virtual proxy and the 3rd virtual proxy only use central node as load balancing node. I am going to add rim node for central virtual proxy and have some questions:

1. Will it impact external users' activity (thru 3rd VP) in any way? I think no but not 100% sure.

2. Currently only central node has SSL certificate implemented. If I add rim node for central virtual proxy, does rim node need to have SSL certificate as well?

Thanks

[RafaelBarrios]

Hi @Bo_PT 

i understand you have something like this

if so, i see two option here:

#1 make all engines available for all virtual proxies

which is great for high availability and making the most of resources

#2 modify standard load balancing rule

if for any reason you want to control the availability and usage of engines, you can keep your current configuration and just make monitoring apps available in all engines by modifying standard load balancing rule as follows

as for your questions:

1. Will it impact external users' activity (thru 3rd VP) in any way? I think no but not 100% sure.

User will be kicked out just for a couple of seconds while proxy gets retarted, and off course, monitoring apps will use second node resources when reloaded or opened.
if you are having resources issues, this can have an impact

2. Currently only central node has SSL certificate implemented. If I add rim node for central virtual proxy, does rim node need to have SSL certificate as well?

Without knowing how your internal and external users access your site, it's difficult to know, but if you're not having any problems right now, you shouldn't with these changes.
Just keep the virtual proxies whitelists in mind and check if you get an insecure site message.

Hope this helps.

Best,