Data I/O Spiking to 100% in Azure SQL DB | Production vs QA










Hi everyone, welcome to another blog of iLearn. 

First of all, it always gives a different "high" when you tackle a performance-related issue. This particular scenario was that the query in our quality assurance environment was running quicker than the query in production. 

Here is how we started the investigation and arrived at the solution.

First of all, started looking at the statistics. 

1. The First check was to see if the auto stats property is set to active or not and it was ON both in Production and QA databases

SELECT name,is_auto_create_stats_on FROM sys.databases WHERE name = 'OurDatabase';  

2. The Second check was to update the stats on the table (in Production) being queried (since we found that the stats were not up to date)

update STATISTICS OurTable;

Didn't do any good, the query was still running slow

3. Finally decided to compare the configuration for Production and QA Database instances and found that the QA is set to 8 vCores and Production to 2vCores and bingo! Found the culprit!!!

Ideally, this should be the first stop as far as the investigation goes. But no issues all part of learning. 

One of the other challenges we faced was how to correlate high Data I/O to vCore. I mean to say how could we say that if vCores are increased Data I/O spikes will be eliminated. Any suggestions on this?


Comments

Popular posts from this blog

Issues Integrating Azure Data Factory with GITHUB | IN spite of admin rights on repository

SQL QUERY NIGHTMARE

Visual Studio Git Error | "terminal prompts disabled"