Question

Huge number of TCP segment retransmissions

Posted December 6, 2019 572 views
CentOSServer OptimizationNetworkingFirewall

I have encountered network issues (documents are printing multiple time by chance) while cups printing from a Digitalocean server (CentOS) to a external printer. Doing a network analysis with Wireshark I saw that by chance packages are retransmitted to the print server and so documents are printed 2 or 3 times.

I have a server at an other cloud provider and with that I don’t have that problems.

A netstat -s for the digital ocean server shows:

netstat -s | grep retr
    650374 segments retransmited
    1 times recovered from packet loss due to fast retransmit
    1 timeouts after reno fast retransmit
    44015 fast retransmits
    5833 forward retransmits
    4373 retransmits in slow start
    1 classic Reno fast retransmits failed
    1428 SACK retransmits failed 

For the other cloud provider it shows

netstat -s | grep retr
    8389033 segments retransmited
    73 times recovered from packet loss due to fast retransmit
    16 timeouts after reno fast retransmit
    54023 fast retransmits
    3285 forward retransmits
    8724 retransmits in slow start
    37 classic Reno fast retransmits failed
    1732 SACK retransmits failed 

The number of retransmissions on the digitalocean server is significant higher than on the other cloud server from a other cloud provider - although the other cloud server has much more load and much more network traffic.

Does anyone have an idea what could be the reason for that?

Thanks for any help.

edited by MattIPv4
Submit an answer

You can type !ref in this text area to quickly search our full set of tutorials, documentation & marketplace offerings and insert the link!