Hi folks,
In Hadoop IPC Client , I found that the maxretries is 45 in case if SocketTimeoutException is got from server(mostly from datanode or namenode). I wonder why a large value 45 is used as max retry time. Possibly there might be a specific reason behind this(because generally the retry value will be 3 or something like that) ? <moreover the max retry for IOexception is only 10>. Clarifications appreciated.. Thanks, Gokul
<<attachment: image001.gif>>