Experiments with MetaTrader 5 at Discovery - page 59
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
My pings are fine on both accounts at Otkritie right now, my average ping is ok. Check the pings to different resources, maybe the problem is on the ISP side.
The ping is fine.
It's strange that the problem was on the demo and the real one. They are different servers...I was beginning to think that developers have gone crazy during trading experiments).
Maybe for me personally imposed glitches ?...))
cancel order #38968458 sell limit 1.00 Si-9.16 at 65888 placed for execution in 65606 ms
Thanks for the reply. That's weird.
I will look into it .
The ping is fine.
It's strange that the problem was on the demo and the real. They are different servers...I was beginning to think that developers have gone crazy during trading experiments).
Maybe for me personally imposed glitches ?...))
cancel order #38968458 sell limit 1.00 Si-9.16 at 65888 placed for execution in 65606 ms
Thanks for the reply. That's weird.
I will look into it .
How do you measure the ping? The servers are different, that's why it's weird. 65 seconds is a cosmic delay, more like a bug. Does the order appear on the chart after 65 seconds as well?
How do you measure the ping? The servers are different, that's why it's weird. 65 seconds is a cosmic delay, more like a bug. On the chart, does the order also appear after 65 seconds?
Yes, also after a minute, and sometimes faster, after 20-30 sec.
But for some reason everything started to "fly" again after automatic downloading of unknown something from the net by some service. But it's probably not related to it.
Maybe because of the fact that at night I "rolled back" Windows 10 to the previous build. ...But, it shouldn't... It's a coincidence.
I don't know what it was.
I'll keep monitoring the situation. Hopefully it won't happen again.
Here's just from a real account:
Setting Buy Limit, removing it, opening on the market, closing on the market within 5ms with a ping of 2ms. This is from a MetaTrader VPS Server in Moscow.
I just came from a real account:
Setting Buy Limit, removing it, opening on the market, closing on the market within 5 ms with a ping of 2 ms. This is from a MetaTrader VPS Server in Moscow.
I'm doing fine now, too - everything is fast. So, it turns out that my delays are not your doing and you have not conducted any experiments.
Does your broker have a technical possibility to throw individual "crutches" on purpose? (for fun, for example).
I hope that these delays were caused by "miracles" from my experiments with Windows 10.
buy limit1.00 RTS-9.16 at91740 placedfor execution in5 ms
I can even try hard arbitrage with this lag!!!
My ping is 60ms and that's enough for my soft scalper strategies)
I'm doing fine now, too - everything is fast. So, it turns out that my delays are not your handiwork and you didn't do any experiments.
Now the exchange is upgrading its infrastructure and recently rolled out new versions of API. There are a number of complaints from traders on different platforms and their own connectors that the execution time is floating and there are obvious slowdowns.
Most likely these are temporary problems and the exchange will fix them. It is not in its interest to have such delays.
Does the broker have the technical ability to throw in crutches on purpose? (for fun, for example).
I hope that these delays were due to the "miracles" of experimenting with Windows 10.
yes with this delay you can even try hard arbitrage!!!
My ping is 60ms and that's enough for my soft scalper strategies)
After the new release and the update of the broker servers, trade execution time and overall latency of trades will be further improved by a couple of milliseconds.
We are putting a lot of effort into tuning all the trading venues and winning every hundred microseconds in the process chain.
No. The gateway is completely direct to the exchange and the broker cannot interfere with it.
So it seems that the order goes first to the server, where it is processed - checked for correctness - and only then goes to the gateway.
https://www.mql5.com/ru/docs/trading/ordersend
"A trade request goes through several stages of verification on the trade server."
That is, before it reaches the exchange (during validation), the broker has the opportunity (theoretically, if provided by the developers) to "mess around" with it. It's not like post-trade control for traders using DMA.
Or, for the exchange version, the function of checking the correctness of the request is performed by the terminal itself (which keeps records of all transactions along with the server and interacts with it asynchronously), so that the server is not burdened with additional calculations? But then there wouldn't be so many states inENUM_ORDER_STATE.
Or am I misunderstanding?
After the new release and the update of the broker servers, deal execution time and overall latencyof trad es will be further improved by a couple of milliseconds.
We are putting a lot of effort into tuning all the trading venues and winning every hundred microseconds in the process chain.
That's cool. But, imho, there's nothing wrong with execution speed. Why increase it further? What other competing platform should we "punish" in terms of speed? Quick is already outclassed in this respect.
Was looking through the plaza documentation and didn't realise, if you've managed to get things up and running with futures - then it should absolutely be easy and time consuming to get the options feature up and running.
Only there are still no options(
Tell me if it's not a secret, of course.
I'll tell you after I update the servers in Discovery.
Most of the time I have about that:
2016.07.07 11:47:11.564 Trades '10644': buy limit 1.00 Si-9.16 at 65057 (65057) tp: 65457 placed for execution in 7 ms
2016.07.07 11:47:11.557 Trades '10644': buy limit 1.00 Si-9.16 at 65057 (65057) tp: 65457
It seems that the request first goes to the server, where it is processed - checked for correctness - and only then goes to the gateway.
https://www.mql5.com/ru/docs/trading/ordersend
"A trade request goes through several stages of validation on the trade server."
The server checks the overall correctness of the order and sends it directly to the gateway.
Thank you. Anyway, the speed of execution (according to the logs) is impressive when compared to QuickBooks.
There is another important question, which can hardly be answered by anyone but you. I would be very grateful if you answer.
1) It's the speed of MarketData. How do I check the relevance of quotes?
The exchange translates in microseconds, which you can obtain through the plaza-
bid_changed t The time of change of the best current bid quote.
ask_changed t The time of change of the best current quote to sell.
And Metatrader - only server time in sec and values of best prices.
If, in addition to the exchange time of price changes, MT would broadcast the time in ms which is periodically synchronized with the time of the exchange server
- the problem would have been solved. Everything would be OK!
The execution speed does not matter, if the trading decision is made blindly by out-of-date quotes. Sometimes they happen (no matter what the reason) to get really sluggish. And you do not want to do any trading when that happens.
//---------------
2) When requesting all ticks via CopyTicks , in MqlTick structure tick.time_msc(time of last price update in milliseconds) gives out time rounded to seconds, which coincides with server time. Same as time// Last price update time. Why it's not what you need - exchange time and transaction time...? You can get it via plaza. And MT5 gets information from it... There is no answer to this question in Service Desk (!)