Lantronix EDS3000PR User Guide - Page 83
TCP Keep Alive, TCP Keep Alive Interval, TCP Keep Alive Probes, TCP User Timeout, AES Encrypt Key,
View all Lantronix EDS3000PR manuals
Add to My Manuals
Save this manual to your list of manuals |
Page 83 highlights
9: Administration Host Field TCP Keep Alive Description Enter the time, in milliseconds, the EDS3000 device waits during a silent TCP connection before the first Keep Alive probe is sent to the remote host in order to keep the TCP connection up during idle transfer periods. Set to 0 to disable TCP Keep Alive, and blank the field to restore the default. TCP Keep Alive Interval TCP Keep Alive Probes TCP User Timeout Enter the time, in milliseconds, to wait between Keep Alive probes in order to keep the TCP connection up during idle transfer periods. Blank the display field to restore the default. Enter the number of TCP Keep Alive probes to send before closing the connection if no response is received. The probes are sent after the initial TCP Keep Alive probe is sent. Valid values are between 1 and 16. Blank the field to restore the default. Specify the amount of time the TCP segments will be retransmitted before the connection is closed. AES Encrypt Key Enter the AES Encrypt Key and select Text or Hexadecimal to indicate format. This configuration field becomes available when the TCP AES or UDP AES protocol is selected. AES Decrypt Key Enter the AES Decrypt Key and select Text or Hexadecimal to indicate format. This configuration field becomes available when the TCP AES or UDP AES protocol is selected. Initial Send Enter the Initial Send character and select either Text or Binary format. This configuration field becomes available when the SSH, TCP, UDP, or UDP AES protocol is selected. Notes: If the keep alive time expires, the user timeout is expired, and there are probes in flight, the connection will be reset. For this reason, it is recommended that if keep alive is used in conjunction with the user timeout, the keep alive timeouts be larger than the user timeout. If it is smaller, what will typically be seen is that the initial probe will be sent, then at the interval where the next probe would normally be sent, the connection will be reset, with no additional probes sent. Also note that in these cases: if the keep alive timer is significantly smaller than the user timeout, probes will continue to be sent for an unreachable host until the user timeout expires. If there is data in flight when the TCP retransmission timeout kicks in, the user timeout is checked as a limiting condition only when the timer expirations would normally be checked during RTO handling. In other words, the user timeout will not be an exact limit; in practice, it will always take somewhat longer for the connection to be closed. The longer the user timeout is, the more likely it will expire between exponentially slower retransmissions, and the connection will not experience an error until the next retransmission timeout is checked. Also note that the user timeout expiration during retransmission returns an error to the application; it does not automatically reset the connection as happens with keep alive timeout. It is up to the application (e.g., tunneling) to close the connection (this happens almost immediately with tunneling). EDS3000 Device Server User Guide 83