144 redo blocks read, 36 data blocks need recovery

Today morning, i saw one of database is not working, i check the alert.log, the database is open.. and smon is not running, i even given shut immedaite, but given a error and then i given startup command, after a while the database is up.... i check the alert.log again

ALTER DATABASE OPEN
Thu Jul 7 09:12:59 2005
Beginning crash recovery of 1 threads
Thu Jul 7 09:12:59 2005
Started redo scan
Thu Jul 7 09:12:59 2005
Completed redo scan
144 redo blocks read, 36 data blocks need recovery
Thu Jul 7 09:13:00 2005
Started recovery at
Thread 1: logseq 28, block 88935, scn 0.0
Thu Jul 7 09:13:00 2005
Recovery of Online Redo Log: Thread 1 Group 1 Seq 28 Reading mem 0
Mem# 0 errs 0: /koala_u01/oradata/VTST31/redo01aVTST31.log
Mem# 1 errs 0: /koala_u06/oradata/VTST31/redo01bVTST31.log
Thu Jul 7 09:13:01 2005
Completed redo application
Thu Jul 7 09:13:01 2005
Ended recovery at
Thread 1: logseq 28, block 89079, scn 1801.2597341987
36 data blocks read, 36 data blocks written, 144 redo blocks read
Crash recovery completed successfully
..

why it happen, let say it happen again, and how to recover the database. and my database is not archive log mode

What you saw was just a plain normal crash recovery (aka instance recovery). What it means is that some changed blocks of commited data has not been written into the datafiles before the database crashed, but as those commited changes were available in redo log file, they were applied to the datafiles during crash recovery. This process made your database consistent again.

This is all done automaticaly during startup proces after any "unclean" shutdown of the database and it doesn't matter wether your database is in archlog mode or not - crash recovery works the same in both modes.