Key Implementation Details
Immediate Instantiation: You do not need to call the custom indicator using Indicators.GetIndicator(). Instantiating the built-in AverageTrueRange and passing its Result directly into SimpleMovingAverage inside Initialize() is significantly faster and uses less memory.
Log Throttling: When a strategy triggers a persistent "Buy" signal (e.g., price stays within an entry zone for 30 seconds), OnTick() will attempt to execute thousands of times. The LogRejection() method restricts the journal output to once per minute, ensuring your broker logs remain readable.
State Reset: _lastRejectionLogTime is reset to DateTime.MinValue upon successful execution, ensuring the next rejection is logged immediately, even if it happens on the very next candle.
Brokers with good VPS partnerships for low latency
Re: Brokers with good VPS partnerships for low latency
Preserve your own money. Scale with the market's money. Exponential growth is the ultimate key.
-
LondonScalper
- Posts: 770
- Joined: Sat Sep 05, 2026 7:54 am
Re: Brokers with good VPS partnerships for low latency
I'd treat any list of broker and VPS pairings as a starting point to test rather than a recommendation, because server locations change with platform and entity. IC Markets is a good example: MT4/MT5 accounts are served from New York and cTrader from London, so an LD4 VPS is right for one account type and adds a transatlantic leg for the other.
The resource point is the one I'd keep. My own flatten drill failed once on a cheap plan, not because of the network, but because the remote session froze while the terminal was loading a second chart set. Since then I test with everything open that I'd run on a normal morning, not an empty desktop.
The Pine zones are fine as a reminder, but in London most of my slippage comes from the minutes after 08:00 and around data. I'd rather mark those by the clock than by a volatility spike detected after the fact.
The resource point is the one I'd keep. My own flatten drill failed once on a cheap plan, not because of the network, but because the remote session froze while the terminal was loading a second chart set. Since then I test with everything open that I'd run on a normal morning, not an empty desktop.
The Pine zones are fine as a reminder, but in London most of my slippage comes from the minutes after 08:00 and around data. I'd rather mark those by the clock than by a volatility spike detected after the fact.