Multiple MFPro Profiles Lag

Hi,

I have recently taken to running MFP on a VPS to cure what I thought was a speed problem with my laptop.

Generally I have one profile per race meeting, as a trigger I use that starts monitoring races at 10 mins before off won't perform correctly if races overlap due to late starts etc.

I found speed still seemed to be an issue and was told by VPS support that " If you run more than 6 profiles with MFP, it will not work correctly, this not due to the computer but due to MFP."

Is this correct and if so is there a way around it?

I have just checked settings and I don't stop monitoring markets, could it be as simple as that or is there something about more than 6 profiles?

Thanks in advance

Scummbags

I'm afraid I don't have the magic bullet for you, but just a bit of info from my perspective.

I run 6 instances of MFP simultaneously, and I find that any more than that produces logging and port number errors (at least that is the message that pops up). You could look at the memory that each instance is using to see if your getting near the limit, as I know that vps only provide a certain amount of memory depending on your subscription package. Bear in mind that the operating system will need a chunk of free memory as well to operate efficiently. I run my 6 instances of mfp with 32gb Ram, and that is stable.

By the way, are you 'Scumbag Trader' of olden days?

Thanks for reply Armani, 32gb Ram sounds very high? I have
WINDOWS SERVER 2012
INTEL XEON CPU E2420 @ 2,3 Ghz
1 Gbit NETWORK CONNECTION
PURE SSD SPACE: 30 GB
MONTHLY TRAFFIC (BANDWIDTH): 1000 GB
NUMBER OF VIRTUAL CPU: 4
DEDICATED DDR3 RAM: 1 Gbit
FULL REMOTE DESKTOP ACCESS

Not sure how that compares and whether more would overcome the problem

By the way I'm not scumbag trader, just chose a name that amuses me

Why do you need a separate PROFILE for each MARKET?!

Maybe profile is the wrong word, I have one instance for each meeting, each running the same trigger. ie. "default", "default copy", "default copy2" etc.

I'm no expert, but from the specs that you posted I would say that the 1gb of Ram may well be the bottleneck. Are you able to up your subscription temporarily and hire more dedicated memory on your vps to see if that improves things?

Quote from: armarni on 12 May 2016, 17:48

I'm no expert, but from the specs that you posted I would say that the 1gb of Ram may well be the bottleneck. Are you able to up your subscription temporarily and hire more dedicated memory on your vps to see if that improves things?

It's 32G, more than MF can use.

Quote from: scummbags on 12 May 2016, 15:00

Maybe profile is the wrong word, I have one instance for each meeting, each running the same trigger. ie. "default", "default copy", "default copy2" etc.

Why? MF can work with all meeting at the same time.

How do you time the speed or efficiency of a trigger?

For arguments sake if I have a trigger with 120 user variables updating every 0.2 seconds monitoring every selection in a 14 runner event that is going to be more memory intensive than a trigger checking two conditions in a two runner event like tennis updated every two seconds.

I know I can check Windows Task Manager but does the connection monitor cover this, am I missing something?

What I'm saying is, is that you can hone something for greater efficiency if you have a way of measuring its execution speed.

To reply to my own last question the way to check a triggers execution speed is to check the trigger log!

Multiple MFPro Profiles Lag

Hello Guest