Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hi..
Could you please help me to know, is it possible to have load balancing on QlikView server,without having cluster environment in place.
Thanks in advance.
Er, I'm not sure I understand your question. Load Balancing in general is the act of distributing the full load between two or more identical instances of a service (e.g. QVS) or a device (e.g. a router) or a platform (e.g. windows or citrix). Those different load balanced instances will act and look like a single larger "virtual" instance than what would be possible with a real single instance.
Load balancing with a single instance is no load balancing. If you mean, "can I load balance two licensed full QVS installations", then yes you could but all load balancer work (e.g. configuration, management and the actual distribution of the load to one of the QlikView Servers) will have to be done by you. In a clustered QVS environment, the web server does the distribution to one of the available cluster nodes according to the ruleset you select in the QMC..
[Edit] I'ld like to add that Load Balancing is usually performed automatically by the hadrware/software itself (just like with a clustered QVS setup that automatically pushes a share of the work to the different instances depending on for example their current load). If you feel like it, you could configure two non-node QVS installations in QMC and publish different documents to one of them. But remember that you'll need separate sets of CALs for each QVS, and the users will have to be routed to a specific QVS depending on which document they would like to open. If you would like to organise load balancing this way, you would need to have two AccessPoints, a network device that fowards users to a particular setup on an arbitrary basis, you would need to publish all your documents on both platforms and you would need twice the amount of CALs as expected... Doesn't sound too attractive, if you ask me.
Er, I'm not sure I understand your question. Load Balancing in general is the act of distributing the full load between two or more identical instances of a service (e.g. QVS) or a device (e.g. a router) or a platform (e.g. windows or citrix). Those different load balanced instances will act and look like a single larger "virtual" instance than what would be possible with a real single instance.
Load balancing with a single instance is no load balancing. If you mean, "can I load balance two licensed full QVS installations", then yes you could but all load balancer work (e.g. configuration, management and the actual distribution of the load to one of the QlikView Servers) will have to be done by you. In a clustered QVS environment, the web server does the distribution to one of the available cluster nodes according to the ruleset you select in the QMC..
[Edit] I'ld like to add that Load Balancing is usually performed automatically by the hadrware/software itself (just like with a clustered QVS setup that automatically pushes a share of the work to the different instances depending on for example their current load). If you feel like it, you could configure two non-node QVS installations in QMC and publish different documents to one of them. But remember that you'll need separate sets of CALs for each QVS, and the users will have to be routed to a specific QVS depending on which document they would like to open. If you would like to organise load balancing this way, you would need to have two AccessPoints, a network device that fowards users to a particular setup on an arbitrary basis, you would need to publish all your documents on both platforms and you would need twice the amount of CALs as expected... Doesn't sound too attractive, if you ask me.
BTW did you read this chapter of the on-line help?
Hi Peter,
Thanks for you response. You understood my question correctly.
From your reply I came to know that load balancing is possible on QlikView server (off course with the help of more than 1 independent QV servers)without having cluster environment in place.
Currently I have 2 independent QlikView Servers (we called it as Primary and secondary) . Each Server has its Access point and CALs. Since we are using these servers since more than 3 years as 2 independent servers, now we do not want to go with clustering them.
However, we would like to implement load balancing on these servers, so that if primary goes down then we can switch over to secondary server.
Could you please help me know what other things we should have to do this and how should we implement this .
Thanks in advance.
AFAIK a network load balancer (a device) is all you need to make this work.
But have you spoken to your Qlik Partner about the possibilities of swapping those two QVS full licenses for a real clustered QVS license with two nodes? To me that would seem to be the better choice.
Have not talked yet. Lets see.
Please shared me doc if you have for load balancing without cluster.
Thanks