First, if you are going to have a QlikView cluster, there will be no "master" or "slave" servers, as they will respond to the same QMC and will act as one server (with two nodes in your case).
Usually, the web server activity is not heavy, so you can have them in the same computer without having significant performance degrade. However, I'd recommend to have them separate QVS and QVWS whenever possible.
If you are going to have a hardware load balancer, you can add another QVWS to the set, so the balancing is even more effective, as there will be two web front ends serving to users.
In regards to IIS: do your users need to use Plugin? Or do you need to use any technical specification in IIS that is not present in QVWS?
Hope that helps.
Using SAN I cannot have a single shared storage of the QVWs and QVDs that is accessible by Publisher and Server.
If I use SAN, it will have to be connected via a NAS node anyhow.
I got this information from the attached document. Please let me know what you think? Please have a look at the attached document.
I am using QlikView 11.2
Clustering QV Servers.pdf 902.1 K
The latest whitepaper is linked to this reply.
The important parts to this are:
QLIKVIEW SERVER FILE STORAGE CONFORMITY
QlikView Server currently only conforms with Windows File Shares. This means that storage must be owned, governed and shared by a Windows instance. The underlying storage type should not matter.
Provided the Storage is surfaced via a Windows Server the storage type is irrelevant.
QlikView cluster requires a windows CIF share. Its common to use a virtual windows files server between your SAN and the QlikView cluster. This provides the clustered QlikView components ( AccessPoint, Qlikview Server or Publisher) with a single UNC path to the shared objects.
For added resilience, many customer use 3rd party clustering, clustering the virtual windows file server (windows virtual head), accessing the SAN.
Hope this helps?