Forum Discussion

Sathishkumar123's avatar
Sathishkumar123
Icon for Nimbostratus rankNimbostratus
Oct 20, 2017

TCP::Collect vs TCP::payload

I have a simple iRule as below

 

when CLIENT_ACCEPTED {TCP :: collect}

 

when CLIENT_DATA { if { [TCP::payload] contains "SomeName/SomeServiceNameHere/"} { pool APoolUsedByMyApplication }

 

My TCP data has 600 bytes. My intended data will be well within first 200 bytes and hence I changed to iRule as below.

 

when CLIENT_ACCEPTED {TCP :: collect 100}

 

when CLIENT_DATA { if { [TCP::payload 250] contains "SomeName/SomeServiceNameHere/"} { pool APoolUsedByMyApplication }

 

I am collecting only 100 bytes via TCP::Collect. But using TCP::Payload 250 and it works fine.

 

Am bit confused here. Do the TCP::Collect and TCP::Payload gets a fresh copy of data? Or Is Data collected in TCP::Collect is reused in TCP::Payload?

 

Please correct my understanding and any comments are helpful

 

Thanks in advance.

 

application delivery
BIG-IP
iRules
  • Kai_Wilke's avatar
    Kai_Wilke
    Icon for MVP rankMVP
    Oct 20, 2017

    Hi Sathishkumar123,

    the 

    [TCP::collect 100]
    command will make sure that you've buffered at least 100 bytes of data before the 
    CLIENT_DATA
    event triggers. If more TCP payload data has been already arrived in the TCP buffer the 
    [TCP::payload]
    command would be able to access this data too. And if less that 100 bytes of data has been already arrived, it will wait until the remaining data has been arrived (Note: This may stall your TCP connection if the client is requiring a response before sending additional data)

    The 

    [TCP::payload 250]
    command will read a maximum of 250bytes from the TCP buffer. If more data is available in the TCP buffer it will accurately trunkate the payload for you. And if less that 250bytes of data are currently in the TCP buffer, it will simply fetch all available data without any complains...

    Cheers, Kai