Re: Database tracking

[Arup Nanda]

Title: RE: Database tracking

Wow!   That definitely is one simple and elegant solution, Raj.   I wonder what you call when the alarm goes off, "heart-broken"? :)   Arup ----- Original Message ----- From: Jamadagni, Rajendra To: Multiple recipients of list ORACLE-L Sent: Wednesday, January 15, 2003 3:14 PM Subject: RE: Database tracking Tom, This is what we do ... We have a schema called "heartbeat" ... owns a table called rhythm (one column time_stamp data type date); 1. Using a perl program, we log on to instance, insert a row in the table with value sysdate. 2. Use dbms_lock.sleep to sleep for 3 seconds. 3. select sysdate-time_stamp from rhythm; 4. rollback; 5. exit from the program. If the difference at stage 3 is > 5 seconds, the web-page goes red, if it is > 6 seconds the alarm goes off. Works fine ... we have this set-up on all our RAC instances ... The query is run from a monitoring server that connects to the database on public network (as everyone else) and also has 2 private networks ... if public network goes away. Raj ______________________________________________________ Rajendra Jamadagni              MIS, ESPN Inc. Rajendra dot Jamadagni at ESPN dot com Any opinion expressed here is personal and doesn't reflect that of ESPN Inc. QOTD: Any clod can have facts, but having an opinion is an art! -----Original Message----- From: Terrian, Tom (Contractor) (DAASC) [mailto:[EMAIL PROTECTED]] Sent: Wednesday, January 15, 2003 1:59 PM To: Multiple recipients of list ORACLE-L Subject: RE: Database tracking I guess I would also like to test out the network response time.  If I run the same SQL from one UNIX box to the production databases at other sites (via sqlnet), I can record total run time and sql statement run time (I assume the difference would be network response time?).  If I keep this information forever then I will know if the databases are slowing down or speeding up.  I could also determine if particular boxes are speeding up or slowing down. The question is, what would be a good SQL statement to test?  Is SQL> select count(*) from dba_tables; as good as another?