We have a 6.3 DT Production server that is "medium" DT sized (recommended RAM is 7GB for this size) but it has 64 GB of RAM.
We have around 10 sys profiles and around 16 concurrent DT clients connected from various users during the day, which is far more that the supported 5 DT client for a medium sized DT server. We are finding that the system can't cope and that when running reports on views, it either takes a very long time or sometimes we get error saying not enough memory. This is backed up with a few Incidents maybe on DT server or frontend server memory is low.
After research I have seen that you can change the frontendserver.ini memory flag which may help the slowness of the DT client or running of views. and the guide states to have a "Xlarge" size to deal with more DT clients and sys profiles. This should speed up the quickness of the DT client, etc.
Could we change this flag from say "medium" to "large" or "xlarge" to see if it helps considering we have 64 GB of RAM. Or is the flag linked to the overall specification of the DT server (i.e. the # CPU cores, etc).
Please advise, if we cant do the above then I guess the next best option is either to increase the spec of the DT server or consider splitting the DT server into 2 and sharing the load?
There are ways to increase the Memory available while keeping some of the other sizing values lower. In this respect, I highly suggest you open a support ticket to talk with them on the specifics of what you want.