﻿<?xml version="1.0" encoding="utf-8"?><rss version="2.0"><channel><title>Ayende @ Rahien</title><link>http://ayende.com</link><description>Ayende @ Rahien</description><copyright>Copyright (C) Ayende Rahien  2004 - 2021 (c) 2026</copyright><ttl>60</ttl><item><title>Gian Maria commented on NHProf new feature: Expensive queries report</title><description>Yes, you are right, in dev environment we can have indicators, just to find if some queries are really really bad, and then refinement should be made on QA or Production machine.
  
  
In my opinion the problem is in the hardware: some devs have machine with 2 gb ram and they test on local sql server, but numbers are really really different when you run the same set of queries on a 2xXeon, 12 GB ram and a bunch of SAS 15k iSCSI disks :). 
  
  
alk.
</description><link>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment5</link><guid>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment5</guid><pubDate>Fri, 23 Jul 2010 12:29:29 GMT</pubDate></item><item><title>Ayende Rahien commented on NHProf new feature: Expensive queries report</title><description>Gian,
  
I agree with you that only on production (or at least QA) you can find the real numbers.
  
But even on non prod systems, that can give you very strong indications about how things are going
</description><link>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment4</link><guid>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment4</guid><pubDate>Thu, 22 Jul 2010 14:47:48 GMT</pubDate></item><item><title>Gian Maria Ricci commented on NHProf new feature: Expensive queries report</title><description>This tend to be a really really nice feature, but I think that query optimization is best done using a profiler in production environment and finding what is happening during real database usage and in real production machine.
  
  
Execution time tends to be different from local sql server, or dev machine respect to production one, but surely this is a nice feature to have.
  
  
NHProfiler really rocks.
  
  
Alk.
  
  
</description><link>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment3</link><guid>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment3</guid><pubDate>Thu, 22 Jul 2010 13:56:35 GMT</pubDate></item><item><title>Ayende Rahien commented on NHProf new feature: Expensive queries report</title><description>Expensive means took the longest time.
  
This is calculated on a single query basis, not across all similar queries.
</description><link>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment2</link><guid>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment2</guid><pubDate>Sun, 18 Jul 2010 08:11:43 GMT</pubDate></item><item><title>Rafal commented on NHProf new feature: Expensive queries report</title><description>What is the definition of expensive query? Long execution time? CPU utilization? IO utilization? Heavy locking? 
  
Also, the total cost of query (TCQ for CIOs) is calculated by multiplying single query cost by number of query executions / unit of time
  
</description><link>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment1</link><guid>http://ayende.com/4554/nhprof-new-feature-expensive-queries-report#comment1</guid><pubDate>Fri, 16 Jul 2010 10:38:22 GMT</pubDate></item></channel></rss>