So we have been using PlayFab for a awhile now while in development and can clearly state that 750 free compute hours isn't enough for a month.... or we are doing something stupid
Here is what we are doing ... 2 devs that's all that is testing and developing a client server app we are working on. We set things up with the F2v2 I think it is ... forget where we got that recomendation but that is the one we have been using. Setting it up for 2 servers per machine so we can go in and out of various sessions without hitting the no more servers message
Typically we hit 700-ish about mid way through the month and I have to belive we are doing something wrong here. I mean massive majority of that time is idle in just 400 real hours we burn through right at 700 compute hourse. Now I get the whole standby system means there can be 2 servers up at a time as we connect and drop out of a given server but any time I check there are 2 on standby when I configured for 1 ... this I assume is because we have 2 servers per 1 VM.
For note that means we are burning at the moment about 1.75 compute hours per 1 real-time hour ... and that is not 1 real time hour of testing tha tis 1 real time hour as in within on 25 hour day we see about 43 compute hours gone.
Why do we have servers up all the time?
So we can do rapid play testing together .... we are widly distributed team working at all crazy hours so a local server doesn't do anything for us, only having a server up 9-5 doesn't do anything for us. We need to be able to pop on and off at will, try some things and move on. This means that we more or less always have a build deployed.
So what are some things we could do to meet the following work requirements
1) Basicly always need a cloud hosted server up for play testing with the team
2) Need to keep it at or under the 750 a month... defently not double that :) ... this isn't about avoiding costs its about understanding how this very low use can equate nearly double time on the clock
3) Need this hosted server to be as close to what would be in production as possible ... so the server allocation aspect and all that not interested in just tossing up a VM on Azure and calling it good enough ... we moved past that when we started up with PlayFab services
Any guidance would be appreciated.