Question
When a connection to the other host fails during sending, it takes a long time before the transfer ends abnormally. Which setting item should I change to shorten the time until error detection?
Answer
The relevant parameters are [Retry Count] and [Retry Interval] in a HULFT Transfer service.
When HULFT Square tries to connect to a peer host and receives a connection failure error from TCP/IP, it will wait for the time set in Retry Interval and retry. This operation is repeated for as many as the number set in Retry Count, and an error is considered to have occurred when a connection cannot be established.
Therefore, the time required to detect a failure in establishing a connection to the other host is calculated as follows.
Time to detect a connection error = T + (T + Retry Interval) * number of Retry Count
*T: Time taken by TCP/IP to return a connection error
The time until TCP/IP connection error returns is the time to detect an error at the TCP/IP level and does not depend on the HULFT Square setting.
If the time until error detection is not reduced by decreasing the Retry Count and Retry Interval settings, the OS and network settings may need to be reviewed and the time until TCP/IP connection error returns may need to be adjusted.
Supplementation
Please refer to the following page for more information on Retry Count and Retry Interval.
HULFT Transfer Service
https://www.hulft.com/help/en-us/HULFTSquare/Content/HULFT_Square/HS_Services/services_HULFT_Transfer.htm
Comments
0 comments
Article is closed for comments.