Oracle9i on Red Hat Linux. SA did an strace of an Oracle server process and reported to me the following: select(1024, [18], [], [], {0, 0}) = 0 (Timeout) [...] oracle 15703 oracle 18u IPv4 456196 TCP nyl01d-8101:1521->199.53.75.14:1318 (ESTABLISHED) Before the reboot, that select system call was repeating thousands of times a minute. (My terminal window has 5000 lines and was completely overrun quickly.) Following the reboot, I would see the select repeating, but not as continously. It would be interrupted by gettimeofday calls and some map operations which said to me that it was actually doing something. Having dug a bit deeper into select man page, the select is basically querying the socket to see if there is anything to be read off and immediately returning -- the final {0,0} value specifies zero timeout. This essentially means the Oracle process is polling the socket to the Windows machine for information coming from its end. I'm not a networking expert--is this normal SQL*Net behavior? Paul Baumgartel CREDIT SUISSE Information Technology Securities Processing Databases Americas One Madison Avenue New York, NY 10010 USA Phone 212.538.1143 paul.baumgartel@xxxxxxxxxxxxxxxxx www.credit-suisse.com ============================================================================== Please access the attached hyperlink for an important electronic communications disclaimer: http://www.credit-suisse.com/legal/en/disclaimer_email_ib.html ==============================================================================