When a bot was "unexpectedly terminated", cTrader shows the bot as still running
01 Aug 2024, 00:38
See screen captures below. I don't know how to reproduce it easily.
But when a bot instance is unexpectedly terminated, cTrader algo tab still shows the bot as “running” and not stopped. Should this be? If a bot instance process is terminated, should the bot show as “still running”? It misleads users into thinking it's doing what it's supposed to, when in actuality it's not doing anything.
I also cannot click on the circled “stop” buttons either - it's unclickable. The only way I've been to fix it is to close down and restart cTrader.
cTrader version 5.0.29
Why is the above EURAUD and below EURCHF bot instances still showing as running when in the log it clearly states the process was unexpectedly terminated?
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
On a separate issue, I'm not receiving email notifications when someone responds to a post in the forums. Are they working?
The issue reported above happened again to me today. AGain the button won't click, I can't stop it. The only way for me to get out of this is to kill cTrader and restart.
When I do that, I do know the “OnStop” even doesn't happen or register, because in my code when a bot instance stops, I have it write information to a text file. All the forex pairs (such as NZDUSD below) don't write anything out when this happens. So whatever terminates the process doesn't cause the “OnStop” even to fire.
I can't figure out exactly how to reproduce it, so am not sending through the full bot code. I do know it happens when I have 26 bot instances running under one instance of cTrader on a VPS using 4GB of memory making all sorts of trades and managing them. Maybe it gets hung when too many ticks come through and it can't handle them? Kind of like when debugging in Visual Studio and if we wait too long it kills the process to safeguard data.
I'm trying to figure out code to send you that easily reproduces this behaviour all the time.
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
On a separate issue, I'm not receiving email notifications when someone responds to a post in the forums. Are they working?
The issue reported above happened again to me today. AGain the button won't click, I can't stop it. The only way for me to get out of this is to kill cTrader and restart.
When I do that, I do know the “OnStop” even doesn't happen or register, because in my code when a bot instance stops, I have it write information to a text file. All the forex pairs (such as NZDUSD below) don't write anything out when this happens. So whatever terminates the process doesn't cause the “OnStop” even to fire.
I can't figure out exactly how to reproduce it, so am not sending through the full bot code. I do know it happens when I have 26 bot instances running under one instance of cTrader on a VPS using 4GB of memory making all sorts of trades and managing them. Maybe it gets hung when too many ticks come through and it can't handle them? Kind of like when debugging in Visual Studio and if we wait too long it kills the process to safeguard data.
I'm trying to figure out code to send you that easily reproduces this behaviour all the time.
Hi firemyst,
If you want to receive notifications for responses, make sure you subscribe to the thread
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
On a separate issue, I'm not receiving email notifications when someone responds to a post in the forums. Are they working?
The issue reported above happened again to me today. AGain the button won't click, I can't stop it. The only way for me to get out of this is to kill cTrader and restart.
When I do that, I do know the “OnStop” even doesn't happen or register, because in my code when a bot instance stops, I have it write information to a text file. All the forex pairs (such as NZDUSD below) don't write anything out when this happens. So whatever terminates the process doesn't cause the “OnStop” even to fire.
I can't figure out exactly how to reproduce it, so am not sending through the full bot code. I do know it happens when I have 26 bot instances running under one instance of cTrader on a VPS using 4GB of memory making all sorts of trades and managing them. Maybe it gets hung when too many ticks come through and it can't handle them? Kind of like when debugging in Visual Studio and if we wait too long it kills the process to safeguard data.
I'm trying to figure out code to send you that easily reproduces this behaviour all the time.
Hi firemyst,
If you want to receive notifications for responses, make sure you subscribe to the thread
Best regards,
Panagiotis
I started this post, thus I'm automatically subscribed. I should have received your response, but nothing. I didn't know until I was browsing today. That's why I was asking. I know about the subscribe feature.
However, I received a notification about your most recent response, so that tells me 1) either you somehow responded differently over the weekend (being an admin and all) that cause me to not receive anything or 2) the service was down for some reason.
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
On a separate issue, I'm not receiving email notifications when someone responds to a post in the forums. Are they working?
The issue reported above happened again to me today. AGain the button won't click, I can't stop it. The only way for me to get out of this is to kill cTrader and restart.
When I do that, I do know the “OnStop” even doesn't happen or register, because in my code when a bot instance stops, I have it write information to a text file. All the forex pairs (such as NZDUSD below) don't write anything out when this happens. So whatever terminates the process doesn't cause the “OnStop” even to fire.
I can't figure out exactly how to reproduce it, so am not sending through the full bot code. I do know it happens when I have 26 bot instances running under one instance of cTrader on a VPS using 4GB of memory making all sorts of trades and managing them. Maybe it gets hung when too many ticks come through and it can't handle them? Kind of like when debugging in Visual Studio and if we wait too long it kills the process to safeguard data.
I'm trying to figure out code to send you that easily reproduces this behaviour all the time.
Hi firemyst,
If you want to receive notifications for responses, make sure you subscribe to the thread
Best regards,
Panagiotis
I started this post, thus I'm automatically subscribed. I should have received your response, but nothing. I didn't know until I was browsing today. That's why I was asking. I know about the subscribe feature.
However, I received a notification about your most recent response, so that tells me 1) either you somehow responded differently over the weekend (being an admin and all) that cause me to not receive anything or 2) the service was down for some reason.
Thank you
Hi firemyst,
We didn't change anything neither we are aware of any issues. Make sure the email did not end in your spam folder.
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
I have managed to create a cBot which caused this issue for me today. It doesn't print out the “Unexpectedly terminated” error, but the bot instances stop responding.
I'll email it through to the “community” email address shortly.
I had to let the bot run over the weekend and noticed certain symbols stopped ticking over today.
My guess now is it's being caused by an async callback not working properly as when I added one into the code, that's when things broke. Eg, I stripped out all the code and started to gradually add things back in to see what would cause it to break.
I'll try and make notes of that in the code.
I discovered this issue by looking at the logs. I hav einformation printed out for every tick – some symbols like AUDUSD were still printing to the log today; others like AUDCHF stopped printing to the log yesterday. Both bot instances were running off the same bot code.
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
Hi @PanagiotisCharalampous :
I've reproduced the issue again. It's in my log files:
I sent through a private YouTube video link to the Community Spotware email address referencing this thread so your team can watch the video and see exactly what's happening.
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
Any updates on this issue?
I replaced the async call with the synchronous code below, and the issue still occurs for me: the bot instance is still marked as “running”, but it's not doing anything/logging, and/or crashes entirely with the error message above:
if (_p.Pips > MaxTrailingStopDistancePips + 2)
{
//I tried replacing the async code with the following synchronous api call in the
//sample code I sent to Spotware, and the issue still occurs.
//Seems 1 instance of cTrader cannot handle 26 running instances of a cbot when using this call
Print("Moving Stop Loss");
ModifyPosition(_p,
(_p.TradeType == TradeType.Buy ? s.Bid - (MaxTrailingStopDistancePips * s.PipSize) : s.Ask + (MaxTrailingStopDistancePips * s.PipSize)),
_p.TakeProfit.GetValueOrDefault(),
_p.HasTrailingStop
);
//ModifyPositionAsync(_p,
// (_p.TradeType == TradeType.Buy ? s.Bid - (MaxTrailingStopDistancePips * s.PipSize) : s.Ask + (MaxTrailingStopDistancePips * s.PipSize)),
// _p.TakeProfit.GetValueOrDefault(),
// _p.HasTrailingStop,
// (TradeResult r) => {
Print("Moved Stop Loss");
_waitForAsyncTPToComplete = false;
// }
// );
//_waitForAsyncTPToComplete = true;
}
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
I have emailed the Community account with details for accessing my VPS so they can see the issue happening first hand. I've explained what to look for in the details of the email.
It seems we have not received this email. Can you please resend it to community@ctrader.com?
Best regards,
Panagiotis
Resent as requested. PLease let me know if you received it. If not, I might have to send separate individual emails with all the info as I wonder if your email system might be blocking it or sending it to spam for some reason.
It seems we have not received this email. Can you please resend it to community@ctrader.com?
Best regards,
Panagiotis
Resent as requested. PLease let me know if you received it. If not, I might have to send separate individual emails with all the info as I wonder if your email system might be blocking it or sending it to spam for some reason.
It seems we have not received this email. Can you please resend it to community@ctrader.com?
Best regards,
Panagiotis
Resent as requested. PLease let me know if you received it. If not, I might have to send separate individual emails with all the info as I wonder if your email system might be blocking it or sending it to spam for some reason.
It seems we have not received this email. Can you please resend it to community@ctrader.com?
Best regards,
Panagiotis
Resent as requested. PLease let me know if you received it. If not, I might have to send separate individual emails with all the info as I wonder if your email system might be blocking it or sending it to spam for some reason.
Just wanting to confirm you've received the messages I've sent? If not, I think Spotware's email system is blocking them for some reason because all my initial messages got through to the Community team - obviously since they asked for the VPS' details to check out the issue.
It seems we have not received this email. Can you please resend it to community@ctrader.com?
Best regards,
Panagiotis
Resent as requested. PLease let me know if you received it. If not, I might have to send separate individual emails with all the info as I wonder if your email system might be blocking it or sending it to spam for some reason.
Just wanting to confirm you've received the messages I've sent? If not, I think Spotware's email system is blocking them for some reason because all my initial messages got through to the Community team - obviously since they asked for the VPS' details to check out the issue.
Thank you
Hi firemyst,
Messages received, the team will check later this week or early next week. Thanks for your patience!
Just wanting to confirm you've received the messages I've sent? If not, I think Spotware's email system is blocking them for some reason because all my initial messages got through to the Community team - obviously since they asked for the VPS' details to check out the issue.
Thank you
Hi firemyst,
Messages received, the team will check later this week or early next week. Thanks for your patience!
Best regards,
Panagiotis
Any updates? HAve they been able to see what's going on?
Just wanting to confirm you've received the messages I've sent? If not, I think Spotware's email system is blocking them for some reason because all my initial messages got through to the Community team - obviously since they asked for the VPS' details to check out the issue.
Thank you
Hi firemyst,
Messages received, the team will check later this week or early next week. Thanks for your patience!
Best regards,
Panagiotis
Any updates? HAve they been able to see what's going on?
Hi firemyst,
Unfortunately our engineers were not able to reproduce the "unexpectedly terminated" on your VPS either. So at the moment there isn't much we can do.
Just wanting to confirm you've received the messages I've sent? If not, I think Spotware's email system is blocking them for some reason because all my initial messages got through to the Community team - obviously since they asked for the VPS' details to check out the issue.
Thank you
Hi firemyst,
Messages received, the team will check later this week or early next week. Thanks for your patience!
Best regards,
Panagiotis
Any updates? HAve they been able to see what's going on?
Hi firemyst,
Unfortunately our engineers were not able to reproduce the "unexpectedly terminated" on your VPS either. So at the moment there isn't much we can do.
Best regards,
Panagiotis
IT WAS RIGHT THERE. ON the screen in the open session as I left it. If they watched the video, they would have seen processes were still indicated as “running” but WERE NOT logging anything on every tick. Eg, they stopped, or “got lost”, or something, but were no longer responding.
Using the exact same source code I emailed them at the Community@Spotware account. I had it open. All they had to do was look and see.
Nobody actually watched the video did they? I see the youtube count didn't go up.
Just wanting to confirm you've received the messages I've sent? If not, I think Spotware's email system is blocking them for some reason because all my initial messages got through to the Community team - obviously since they asked for the VPS' details to check out the issue.
Thank you
Hi firemyst,
Messages received, the team will check later this week or early next week. Thanks for your patience!
Best regards,
Panagiotis
Any updates? HAve they been able to see what's going on?
Hi firemyst,
Unfortunately our engineers were not able to reproduce the "unexpectedly terminated" on your VPS either. So at the moment there isn't much we can do.
Best regards,
Panagiotis
This is from the VPS as it is right now. I've left these windows open.
I can't make it any more obvious what the problem is.
I did not write that this does not happen, I wrote that we are not able to reproduce the "unexpectedly terminated" message on our machines on your VPS either. If we cannot reproduce it, then we cannot easily spot the problem. Also we cannot spend unlimited hours waiting for one user's problem to happen again. If you are able to isolate exact steps that will allow us to consistently reproduce this behavior, we are happy to have another look. Else it is a low priority issue and we will fix it if we stumble on it in another occasion.
I did not write that this does not happen, I wrote that we are not able to reproduce the "unexpectedly terminated" message on our machines on your VPS either. If we cannot reproduce it, then we cannot easily spot the problem. Also we cannot spend unlimited hours waiting for one user's problem to happen again. If you are able to isolate exact steps that will allow us to consistently reproduce this behavior, we are happy to have another look. Else it is a low priority issue and we will fix it if we stumble on it in another occasion.
Best regards,
Panagiotis
I explained how to do it in the video that apparently nobody from Spotware is willing to watch from start to finish.
I also sent them this email, dated August 20th, that I was acknowledged because they responded to it:
I did not write that this does not happen, I wrote that we are not able to reproduce the "unexpectedly terminated" message on our machines on your VPS either. If we cannot reproduce it, then we cannot easily spot the problem. Also we cannot spend unlimited hours waiting for one user's problem to happen again. If you are able to isolate exact steps that will allow us to consistently reproduce this behavior, we are happy to have another look. Else it is a low priority issue and we will fix it if we stumble on it in another occasion.
Best regards,
Panagiotis
I explained how to do it in the video that apparently nobody from Spotware is willing to watch from start to finish.
I also sent them this email, dated August 20th, that I was acknowledged because they responded to it:
Proof of the acknowledgment of having received the email. THe “unexpectedly terminated” doesn't always happen, but what DOES always happen is the bots will “lose” their connection to cTrader. I've left the windows open in the VPS showing this. AUDCAD pair. cTrader says the bot is still running, but it's obviously not doing anything because it's no longer logging anything in the “OnTick” event. This is easily reproducible in the steps I've outlined in the above email screen capture. It happened for your team as well – just look at the logs from when they started the bots. They should see several bots are still running, but cTrader “lose” them because they are no longer responding/logging.
I did not write that this does not happen, I wrote that we are not able to reproduce the "unexpectedly terminated" message on our machines on your VPS either. If we cannot reproduce it, then we cannot easily spot the problem. Also we cannot spend unlimited hours waiting for one user's problem to happen again. If you are able to isolate exact steps that will allow us to consistently reproduce this behavior, we are happy to have another look. Else it is a low priority issue and we will fix it if we stumble on it in another occasion.
Best regards,
Panagiotis
Any updates on this issue? It's affecting us on our VPS as well. cTrader keeps “losing touch” with running bot instances as described in this thread. We discovered this two weeks ago when we were wondering why our bots weren't taking trades they were supposed to. They just “stopped” logging anything, but cTrader still showed them as running.
Any updates on this issue? It's affecting us on our VPS as well. cTrader keeps “losing touch” with running bot instances as described in this thread. We discovered this two weeks ago when we were wondering why our bots weren't taking trades they were supposed to. They just “stopped” logging anything, but cTrader still showed them as running.
It's frustrating that now there's at least 2 users reporting this issue, and there still doesn't seem to be any movement on it even though I:
sent them emails (Screen captures above in this thread where they acknowledge receipt of the emails) in August on what the issue is and how to reproduce
provided them the exact source code I was using
provided them full access to my VPS with the bots running where the issue outlined in the emails happens all the time
provided them a YouTube video link demonstrating the issue and how I reproduce it
PanagiotisCharalampous
01 Aug 2024, 08:02
Hi firemyst,
Can we have the cBot that reproduces this problem?
Best regards,
Panagiotis
@PanagiotisCharalampous