Following up on this it seems my existing Python 2.7 data recorder is connecting ok and getting lots of Lighstreamer PROBE messages but never any data. My prototype new recorder (Python 3) works just fine.
Clearly I should just finish my new recorder and use that.
But I wonder if something has changed on the server side over the weekend to cause this?
Hi,
From the market open last night you will no longer be able to subscribe to the VIX and to the Russel 2000 index. As per what Mark said this is due to the agreement we have with the exchanges which explicitly says to not make pricing available via the API.
I do apologise for the inconvenience.
Thanks,
Chris
Confirmed that the Russell was the problematic symbol. The Python Lightstreamer client I had forked for my own use was throwing away the HTTP response from a subscription attempt and so never saw this:
# 2016-06-13 21:29:39.710048 DBG b'ERROR\r\n-1\r\nInstrument not redistributable\r\n'
Is now being logged so hopefully in the event of something similar it'll be easier to diagnose.
Is there a channel of notifications IG has for changes like this? Twitter etc.?
Hi,
The change made so that the VIX and Russell were no longer able to be used on our API was something that applied in a very short time frame for which there was little warning.
Should we hear of any changes of this nature we will gladly add this to the Labs community.
I do hope that helps and do let me know if there is anything else I can do.
Thanks
Chris
I am getting an KeyError: 'cst' when trying to connect. This worked correctly on Friday and previously but failed this afternoon. My code has not changed for weeks. Is this a known problem and if so when will it be fixed?
Just checked and everything was working on Friday up to about 9pm and was not working from about 3am today (18th July). I have tried a demo and live account and they are behaving the same way so either there is a system error or something has changed in the API which is causing me problems. Has there been an authentication change? Everything looks okay from the web / mobile interfaces so it looks as if it just the API that I am having problems with.
Following up on this it seems my existing Python 2.7 data recorder is connecting ok and getting lots of Lighstreamer PROBE messages but never any data. My prototype new recorder (Python 3) works just fine.
Clearly I should just finish my new recorder and use that.
But I wonder if something has changed on the server side over the weekend to cause this?
Seems it's a problem where the subscription set includes MARKET:TM.D.RUS2000.DAILY.IP
Can any other lightstreamer users confirm this?
Yes am having the same problem, ringing the helpdesk now
i'm using the api to download data into an excel sheet. it works fine. when i add
TM.D.RUS2000.DAILY.IP
and call
sHostName = Environ$("computername")
Application.WorksheetFunction.RTD "IG.api.excel.RTD.IGApiRTDServer", sHostName, "Refresh"
my sheet crashes
Just spoke with technical support. IG are not distributing price data for the russel 2000 going forward due to a contractual issue.
Confirmed that the Russell was the problematic symbol. The Python Lightstreamer client I had forked for my own use was throwing away the HTTP response from a subscription attempt and so never saw this:
# 2016-06-13 21:29:39.710048 DBG b'ERROR\r\n-1\r\nInstrument not redistributable\r\n'
Is now being logged so hopefully in the event of something similar it'll be easier to diagnose.
Is there a channel of notifications IG has for changes like this? Twitter etc.?
Thanks for the confirmation. Two questions:
1) Would it be possible for IG to proactively notify API users of changes like this?
2) Are VIX and Russell special cases or is there a risk that we will see more real time price data gradually being removed from the stream?
I am getting an KeyError: 'cst' when trying to connect. This worked correctly on Friday and previously but failed this afternoon. My code has not changed for weeks. Is this a known problem and if so when will it be fixed?
Philip
Just checked and everything was working on Friday up to about 9pm and was not working from about 3am today (18th July). I have tried a demo and live account and they are behaving the same way so either there is a system error or something has changed in the API which is causing me problems. Has there been an authentication change? Everything looks okay from the web / mobile interfaces so it looks as if it just the API that I am having problems with.
I had no data, until logged into dealer to accept new t&c, then data flow restarted. I am based in the UK.
I agreed yesterday evening and all is well this morning. It would have been good if we had been told this in advance though. Thanks xnox.