Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    text
    copied!<p>Memory IMHO is the most important for <code>Oracle</code> performance.</p> <p>Running multiple databases implies keeping multiple sets of cached <code>SQL</code> and <code>PL/SQL</code>, multiple data cache for system tables, etc.</p> <p>If you just need to keep the data separate, you may create different <code>TABLESPACES</code> for different users. You'll still have to share <code>LOGFILES</code> though.</p> <p>Disabling extra features like <code>Data Mining</code> won't help you much, as they do not consume memory when not being used.</p> <p>You'll certainly need to lower any memory values, but it's hard to tell which you should keep and which you need to lower without seeing you database design.</p> <p>As a <strong>very imprecise</strong> rule of thumb, if you have <code>OLTP</code> database, that is little tables and high level of concurrency, you probably should sacrifice <code>sort_area_size</code> and <code>hash_area_size</code>, but keep <code>db_block_buffers</code> as high as possible.</p> <p>If you have large tables using <code>HASH JOINS</code> and <code>MERGE JOINS</code>, you'll need <code>sort_area_size</code> and <code>hash_area_size</code> for efficient joins and ordering, but you can decrease <code>db_block_buffers</code>, as you'll not be able to cache these tables anyway.</p> <p><code>Linux</code> and <code>Windows</code> do not differ much in terms of <code>Oracle</code> performance. <code>Linux</code>, though, graces <code>LOCK_SGA</code>. <code>Windows</code> tries to do it too, but can swap out in tough memory conditions.</p>
 

Querying!

 
Guidance

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload