This is about a datawarehouse ETL job. It's scheduled via crontab and runs once a month. Before a good run, I move the affected tablespaces in read-write by running a script manually, not via crontab like the rest of the ETL. When the load finishes I move them back to read-only mode, same way as at the beginning. Last week the ETL just run second time due to a crontab schedule mistake. Obviously, in this case the tablespace alter to read-write didnt happen, saving the most time consuming part of the ETL, the table loads. The sql*loader part is protected by a sentinel file. The truncate table doesnt happen on a read-only tablespace. So I am left with the indexes - how do I protect them? Rebuilding the indexes is easy and takes about 4h. The problem is the data warehouse slows to unusable until I do it so I want to foolproof it. Thanks, -Carmen Rusu >>> "Khedr, Waleed" <Waleed.Khedr@xxxxxxx> 3/3/2005 11:03:17 AM >>> You need to configure accounts and security differently. Don't give owner id to developers. Waleed -----Original Message----- From: Carmen Rusu [mailto:carmen.rusu@xxxxxxxxxxxxxxx]=20 Sent: Thursday, March 03, 2005 11:03 AM To: oracle-l@xxxxxxxxxxxxx Cc: Carmen Rusu Subject: how can you protect read-only indexes? Oracle EE 9.2.0.4 64 bit SunOS 5.8 64bit data warehouse db ~100gb right now, growing every month no partitioning yet Just verified that you can drop an index when its underlying tablespace is in read-only mode. It happened when an ETL job ran second time, by mistake. The corresponding tables, also on read-only tablespaces, survived ok the truncate op. So, what can I do to foolproof my ETL, so that the indexes are not dropped by mistake next time?=20 Thanks, -Carmen Rusu Oracle DBA -- //www.freelists.org/webpage/oracle-l -- //www.freelists.org/webpage/oracle-l -- //www.freelists.org/webpage/oracle-l