>Everything at this place is on VM - nothing I can do about it as it’s >contractor managed. (In fact they seem to be able to change >the “hardware" without me even taking the server down - that’s something I >never thought I’d see and am too old now to bother >speculating how they do it :-) !).
Yeah, it’s almost magic, that V-Motion thing! >Would increasing the CPU speed help with the trigger execution or is it all >bandwith ? It’s been a few years (3-4) since I had a 4D Server running a VM, but at that time the best way we had to optimize the hardware was to have the VM “Priority” for your app be set as high as possible and to dedicate CPU and RAM to it. VM’s are inherently ‘smart’ and will change the performance level of a given app based on how busy it is. If it doesn’t look like it needs RAM its allocation will be reduced. Likewise for CPU. And the opposite if resources are in demand, the VM will give you app more RAM and CPU. However, it takes time for a VM to react and there is overhead in moving/allocating, so it can effect your apps performance. We found that we needed to set the VM to dedicate RAM and CPU and never reduce it. I don’t recall what the feature was call that does the VM management. That’s the first thing I’d do to get a baseline then start looking at 4D trigger optimizations. Oh yeah, install 4D’s Info Report Component, if you haven’t already. It will tell you, among other things, how your 4D cache is being managed. HTH, Tom Benedict > On Feb 19, 2020, at 13:36, Peter Jakobsson via 4D_Tech <[email protected]> > wrote: > > …that is the question ! > > I have a site that I just launched - usual moderate-trauma of initial days > but basically holding up so far and improving by the day. > > I now need to try and optimise it a bit as it’s still slightly sluggish and > was wondering what people were running in terms of hardware. Everything at > this place is on VM - nothing I can do about it as it’s contractor managed. > (In fact they seem to be able to change the “hardware" without me even taking > the server down - that’s something I never thought I’d see and am too old now > to bother speculating how they do it :-) !). > > I have quite a lot of trigger-code for doing journal posting and so on so I > suppose that does a fair bit of “slowing” running in the co-operative thread. > (The support guy at the contractor noticed it remotely…”hey your DB is all > crammed into a single core !”. ) > > Specs are: > > 4D Server v15. > 8 users. > > Intel (R) Xeon (R) CPU, E5 2620 v4, 2.1 GHz > SSDs (Quite fast ones but not sure of the exact spec) > Windows 10 Server 64 Bit > > Would increasing the CPU speed help with the trigger execution or is it all > bandwith ? > > I’m also going to do some logging-mode execution as I’m sure there are > sequential searches going on in a few places as I usually try to keep indexes > light to start with and wait till I see where I need them. I have a parallel > server running on another machine with my partner license so I can try stuff > out there. > > Any tips and/or experiences appreciated ! > > Regards > > Peter ********************************************************************** 4D Internet Users Group (4D iNUG) Archive: http://lists.4d.com/archives.html Options: https://lists.4d.com/mailman/options/4d_tech Unsub: mailto:[email protected] **********************************************************************
