Solaris

Hi,

Recently we faced awkward situation where we cannot connect to Oracle DB that hosted in Solaris server intermittently and it timeout eventually. The server seems fine. We can putty to server like normal but when try do prstat, it takes time and come out Please Wait.. for a while. When we go and check the server log there was an error say "accept:Software caused connection abort, last message repeated 1 time, getpeername: transport endpoint is not connected".

I also got check the tcp listendrop value using "netstat -sP tcp | grep ListenDrop" and the output as below;

tcpListenDrop = 61080 tcpListenDropQ0 = 0.

Can someone advised whether there should any network contention and how to further check on it. I also will appreciate whether oracle listener able to handle concurrent connection

Ensure that /etc/Inet/hosts and /etc/inet/inodes are linked.
Ensure that /etc/nsswitch.conf has only active name services, and that files is always first.