Number of Ticks normally received per second

RP3436
I know the subject has been discussed before in the forum but I still have some confusion. Therefore I request if this query is considered and replied in simple words without much jargon. Following are my observations and questions:
1. I subscribed to 4 tokens : 2 from NSE and 2 from MCX and collected ticks for 13+ minutes which is approx. 780 seconds. Number of ticks received and collected were : 2836 ticks. This comes to around 1 tick per token per second on an average. Now my questions:
2. Is my observation as expected? If I increase number of tokens to 10 or more - what will be the impact in terms of performance , if I am likely to lose few ticks? If I am on a 40+ MBPS internet line with windows 10 , intel i5 dual core , 4 GB RAM machine (without any other application running) , what is the recommended number of tokens to be subscribed without any significant performance hit. Not using any DB for storing .

This will help me in deciding my scheme and hardware requirements. Pl help.
  • sujith
    Your observation is correct. You can expect up to 1 tick per second for a highly liquid stock. You can subscribe for up to 3000 tokens per connection and with one api_key you may open up to 3 websocket connections.
    The number of subscriptions will not affect any performance from the Kite Ticker end. You need to make sure that your solution should just keep accepting the ticks without blocking the main thread.
  • RP3436
    Great answer. Thanks a lot. Just one more- when you say - "your solution..." - I think the solution you are referring to is not only about the python code but also about the network capability/memory etc. as well . However if the main thread is kept unblocked from other processing activities then kite ticker i.e. your end is not a limitation. Having understood this , my second question relates to subsequent processing. I understand your responsibility is limited to kite ticker end, but if you can OR anyone else, can share some views on possible hardware and network specifications(at user end) like RAM / CPUs/internet speed etc ,the same will be of help. Maximum tokens may be 10-15. Efficient coding is not an issue and has been taken care.
    Regards
Sign In or Register to comment.