I have an app which shows a table.
When I want to access this app in multinode environment through the hub - it's just not working (when I clik on it just the "bubbles" appear on the screen and they are poping there forever). As soon as I removed the other node from QMC I was able to run the app through hub.
Has anybody experienced something like this before ? Any idea ?
There can be many possible reason for such behavior.
A couple of things to check-
Please elaborate little more about the environment and configuration. If possible, please paste log entries where you see the problem is.
I have installed 4 services - engine, repository, scheduler and proxy. Unticked central node. And added in QMC on the master server as a node. The certificates are okay, code came up and entered on the salve browser. The purpose of the node I choose both.
When I want to reach the hub of the slave node, I got an error message: "No available qlik sense engine was found. refresh. or contact administrator...
I only have the original sync rules - so sync everything between the slave and master, and sync everything between master and slave except governance apps.
I can ping the slave machine from the master. So the network is fine, firewall switched off.
There's no proxy between them... if you curious about the central qlik proxy settings they are both the same, the default settings.
Hope it give some help.
I have little confusion here and would like to ask you. Is it the hub or app you are trying to access when you get the error message about No Qlik Sense Engine found?
Can you reach to the hub of the Slave node?
Sorry for making you confused
So I have had 2 situations:
1., Sometimes - mostly after restart the Qlik services - I can access the hub with the master server's URL (even if the slave node added to the central proxy's load balance). I can see the apps, but I can't run them. If the slave is added to the central proxy's load balance. Sometimes after I add the slave node to central proxy's load balance option and I try to access the hub via the master server's URL I get an error message about Connection lost. Make sure Qlik Sense is running properly...
2., When I want to access the hub via the slave server's URL I get the error message about the no available Qlik Sense engine was found...
Hope it helps to understand what's going on here.
Looking forward to your answer.
It seems like there is something wrong with slave node. Everything is fine on central node.
When you see Hub on Central node you see the app. You click on the app and you get the error. That is because you are directed to 2nd load balancing node which is this problematic node. There can be many reasons for this error but what I think most probable cause is either the app is not synced to Consumer Node (slave node) or there is something wrong in the authentication method on 2nd Node.
As you said that when you try to access on slave node URL you get the error, so the error message always comes from 2nd Node.
If you can not access hub on 2nd node then there is something wrong with authentication or in setup for the 2nd node. And if you can access hub on 2nd node but not the app, then there is something wrong with the app on 2nd Node.
I would request you to check Authentication method for 2nd Node proxy and on Central Node proxy as well.
Make sure the nodes are in same domain.
If you see no problem until this point, I would request you to create a new app on Central node, wait for 15-30 seconds and then move to second node to check if you can access there.
Please let me know your further findings. If you still see the problem, please check in proxy logs on Central and Slave. If you see that you managed to reach to slave node and request was not fulfilled by Engine, please paste the error logs from Engine on Slave node.
Thanks for your help. What I did was, deleted the app from the central node. Added the slave node to the load balancing, published the app again. Now the situation has been turned... lol ... sounds strange. Now I can access the hub with the slave node's URL without any problem also able to run the app (loads in fine). The problem now is the central node, now every time I want to access the hub on the central node I get the authentication window popping up and authenticates, but after get an error message - Connection lost. Make sure Qlik Sense is running properly. All is happening now on the central node. I repeat now the central node is the one which is not working and the slave node is fine. It's weird.
The nodes are in the same domain.
If you told me which file I should paste for you (there are 3 or 4 different logs, like proxy system, proxy performance etc.)
Good to know that you do not see the earlier problem anymore.
According to your description of the new problem, I think there is something wrong with proxy and load balancing setup.
You're having two proxies (Central and Node2). If you have both as load balancing node on Central proxy, you should be able to use both nodes without any problem. But if you are trying to access Central node using the proxy of 2nd Node and do not have both as load balancing nodes on that proxy, you will only get access to 2nd node (the node itself) and central would not be be accessible.
If you have seen everything is correct until this point, I would request you to follow instructions as mentioned in MultiNode Configuration Document for Load Balancing Configuration. I would like to know which version are you using of Qlik Sense.
A little instructions if you find them useful-
To make sure that you are accessing central node, remove 2nd node from load balancing node under Central proxy settings. Save it. Restart Message might occur. Ok to Restart. Restart of the service would occur. Refresh the QMC Proxy page. First try to see if you can access Central node hub using the machine and domain name in URL which should be like this https://nodename.domainname.com/hub
That should work and if it works it confirms that there is nothing wrong on Central node as you do not have any other node as load balancing node.
If this doesn't work, there is something wrong in Authentication method. However I do not see any reason for this not to work.
If it works, now is the time to add 2nd node as load balancing node. I hope the 2nd node is added to the cluster already as you have mentioned it assuming that you have not changed default settings for ports and all.
Go the central proxy settings in QMC and add 2nd node as Load balancing node. Save. Restart message. Click Ok. Refresh. ( I assume you do no have any virtual proxies)
To make sure that you have same settings under 2nd node proxy, you have to follow the same process if you want to access both nodes using both URLs.
Hope it will help you out.
I could manage to get it working. I just dropped the slave node, reinstalled Qlik Sense and add it again as a node to the central. Now it works. I ha a chat with the support and the answer was could be certificate issue.
At the moment is working.
Thanks for everyone the answers and the help.