[Spread-users] reason for lots of retrans?

[Jan Rickes]

hi *,

we did some tests with spflooder and spmonitor showed this:

Status at xxx V 3.16. 2 (state 1, gstate 1) after 51718 seconds :
Membership  :  2  procs in 1 segments, leader is yyy
rounds   :   96760      tok_hurry :   25416     memb change:       2
sent pack: 1110096      recv pack :      10     retrans    : 1092638
u retrans: 1092638      s retrans :       0     b retrans  :       0


from the manual i know that retrans = u retrans + s retrans + b retrans.
what's the meaning of these retrans values?


the daemons are in the same segment and are conntected to the same switch.
what can cause this 98% of retrans?


thanks in advance
-jan