http://coskan.wordpress.com/2009/05/27/what-if-vendors-code-was-wrong/ What I learned about Oracle Mon, 06 May 2013 17:37:46 +0000 hourly 1 http://wordpress.com/ http://coskan.wordpress.com/2009/05/27/what-if-vendors-code-was-wrong/#comment-3524 Wed, 27 May 2009 19:02:58 +0000 http://coskan.wordpress.com/?p=368#comment-3524 I had a similar thing happen with an upgrade to a vendor supplied product I was running once. In the test instance the upgrade was taking about 2 days to run. I did some looking, added some indexes, gathered some stats and it ran in 5 hours.

]]> http://coskan.wordpress.com/2009/05/27/what-if-vendors-code-was-wrong/#comment-3522 Wed, 27 May 2009 14:13:36 +0000 http://coskan.wordpress.com/?p=368#comment-3522 I totally agree with your observations especially with a common belief that slow run on development is fine, and this won’t happen in production because it is a bigger box. In our environment we do have policy to review codes if they are reviewable like SQL scripts before execution. Surprisingly, many vendors’ SQL scripts are written with potential issues (no index or index tablespace, cryptic table names like B1004 or H629, etc.) which can easily be remediated. And monitoring the executions is definitely a must. We often repeat the executions on development to ensure the consistent results. Your ending of “Don’t trust word of anybody without any proof” is very true. Thanks.

]]>