Our VPS is consistently at +900mb/s. Since you're doing trading Forex, send us an email to support@ninjamobiletrader.com to let us know if you want it located in Chicago, NYC, or London. We also offer 20% monthly discount for anyone switching from a different provider.
MT5 reads <1ms ping on the platform itself (it's chicago location vps), but the packet loss was like 3% which was weird. RIght, that's what I meant Ok, I'll try yours. I'll email you and need that 20% discount since I'm switching from forexvps. Thanks.
Your situation/problem is not related to the hardware. Your code is probably not optimal and is not covering all possible cases. For applications related to trading, timing is everything. 2 different instances will not process everything at the same exact time, thus resulting in different outcome if your code is weak.
How do you explain a bot that worked for couple of months, where similar trading situation occured that required the b/e function, executed perfectly. And then when I added this new instance, the very next day the b/e fucntion didnt work properly on my instance, but does on the other (including another instance on a separate vps). The only variable was adding the new instance on my vps which used the same magic number. I also noticed after tests today, it is "slower" than my other vps which indicates this company have inconsistent quality. Like seriously, the bot either works all the time or it doesnt. Ive checked the history and traced all the execution line by line, the only discrepancy is that target was placed incorrectly this one time. When I looked by past couple of months, similar conditions occurred, and b/e worked just fine then. So it's not the bot, it either means the vps/connection sucks that cause data loss or maybe the platform sucks like hell for futures trading. I don't know what it is and it's pissing me off.
IS MY TRADING STRATEGY SO HARD TO IMPLEMENT? IT'S A SIMPLE BREAKOUT STRATEGY WITH STOP AND REVERSE FUNCTIONS AND B/E FEATURE ADDED!!! I'VE HAD MULTIPLE DEVS TRY THIS AND I STILL GET ERRORS. YOU THINK THAT THESE DEVS ARE THAT INCOMPETENT THAT THEY CAN'T BUILD THIS STRATEGY? THESE DEVS HAD OVER 100 REVIEWS WITH AVERAGE 4.5 STARS OR HIGHER ERRORS HAPPENED ON MT5 AND QUANTOWER. SO I DON'T KNOW, MAYBE I HAVE A GOD DAMN HOLY GRAIL AND THE FREAKING MARKET CAN'T TAKE IT LOL
There is a possibility you are looking for causes in the wrong places... I found this article on MT5 synchronization with an AMP/CQG setup. To me, it deserves a deeper look, as it sounds like it may be applicable to your problem. https://faq.ampfutures.com/hc/en-us...xplanation-of-Synchronization-in-MetaTrader-5
I dont think so. I specifically had the dev code it so that the EA doesn use any info from the terminal (commission, sync) and calculate pnl internally separately. Cause i knew these things would mess up the calculation if I relied on it, especially since AMP never updated their commission rate since the last CME update. See the thing is, no matter what reason people give me in regards to the code, it doesn't explain how there are discrepancies between instances. It can only be either the shitty vps or the shitty platform (which i can't narrow down the exact mechanics). But i'll take care of the former. I'm going to use ninja vps and use a different magic number across all instances. Let's see what happens.
The article mentions one specific routine regarding the data feed. Honestly, I would not have thought the data feed, and presumably one particular data feed, would be the cause for the problem in the article. What other checks and validations are being done on the live account behind the MT5 screen running multiple instances?? Don't be so quick to say "I don't think so". Keep us updated.
If you think it's a magic number interaction, try making the magic numbers different for each instance.