Very often, significant performance benefits can be obtained by using some very basic knowledge of the application, its data and business rules. Sometimes even less than that: even if you are not familiar with the application logic at all, you can still use common sense to make some reasonable guesses that would get you a long way in improving query’s performance. Here is an example (based on an actual query that I had to tune today).
Database query tuning is mostly about getting better plans. Mostly, but not always. Sometimes, the problem has nothing to do with the plan, and you might need to get a bit creative to find a solution. In this recent case a query was showing a decent performance when running from SQL Developer, but it took about 5 times longer to complete when running from R. The plan was the same, so I knew that it was irrelevant. The R session wasn’t showing as active most of the time, so it was fairly clear that the problem was fetching data — i.e. it was fetching too few rows at a time which lead to a large number of roundtrips, and consequently, high waits on “idle” event “SQL*Net message from client”.
On one of the databases I’m looking after (220.127.116.11, Solaris, non-RAC), several different INSERT statements (all into tablespaces with manually managed segments) suffer from occasional hiccups. The symptoms are always the same: in one of the sessions, the INSERT gets stuck doing lots of single-block I/O against one of the indexes on the inserted table, and if other sessions are running similar INSERTs, they hang on enq: TX – index contention. The situation can last just a few seconds, but sometimes it’s much longer than this (several minutes), in which case the impact on the application is quite serious.
Imagine the following situation: you are supporting an application with many different components and a busy release cycle. One a Monday morning you find that quite a few processes in the database now run slower. Very soon, you find out that the slowdown is due to increased CPU time, but where to move from there? There is no evidence that CPU is too stressed, causing CPU queuing. You cannot isolate the problem to any specific PL/SQL procedure or SQL id, and there seems to be no relationship between affected SQL statements. You also check the changes that went in the last weekend — there are quite a few of them, but none seems to be particularly relevant. So what do you do?
When a query contains a regular or inline view, there are 3 basic strategies for the optimizer to choose from:
1) merge the view (no “VIEW” operation in the plan)
2) instantiate the view as the whole and join it to the rest of the query (the plan shows a VIEW “operation”)
3) push join predicates inside the view (the plan shows “VIEW PUSHED PREDICATE”).
In my previous post I showed an example of how a query’s performance can be improved using the waste minimization technique. My focus was primarily on identifying and enforcing the correct plan, but I received some questions regarding the root cause of the problem: why the optimizer came up with a wrong join order? It’s a very interesting question, and it deserves a separate post so that it could be explored in detail.
Today I’d like to share another tuning example from a recent case at work, which in my opinion is good for illustrating typical steps involved in SQL optimization process.
I was handed a poorly performing query with a relatively verbose text, so I will only give the general structure here (it will also prevent me from accidentally disclosing some sensitive information from that application):