Tune, Monitor and Analyze the Health of Your Centricity SQL Server

Presented by Steve Stedman and Derrick Bovenkamp

About Us

Steve Stedman

• 26 years of SQL Server experience

• Recent experience with Centricity over the last 2 years.

• Blog regularly at http://SteveStedman.com

• Founder of the Database Corruption Challenge.

• Creator of Database Health Monitor http://DatabaseHealth.com.

• Freelance Database Consultant (SQL Server). Stedman Solutions, SQL Data Partners.

• Volunteer Firefighter / EMT with local fire department.

Derrick Bovenkamp

• Senior Systems Administrator at Family Care Network

• 11 Years of experience with Centricity

• Experience with VMware, Citrix, Veeam

• CHUG Board Member

• Teach Health Information Technology at a local Technical College

Family Care Network

• 104 Providers across 13 locations across two counties in Northwest Washington State.

• Locally owned and physician led.

• On PM since 2002 and EMR since 2005.

• Combined on CPS in 2013.

• CPS Database is 800GB.

Introduction

• Performance tuning can save time and money.• Doctors time, Clinical, Front and Back Office Staff time,

Patients time.

• Get more out of your hardware without having to upgrade.

• A stable and reliable database can save time and money.

• System checks and best practices can prevent outages and data loss.

You can download all the slides at http://SteveStedman.com/Chug2016

Baseline Monitoring

Monitoring Tools - Database Health Monitor

Wait StatisticsEstablish a baseline and see changes over time.

Baseline Monitoring

VM and Shared Storage Implications

ESX Top

http://www.running-system.com/vsphere-6-esxtop-quick-overview-for-troubleshooting/

Baseline Monitoring - Memory

Page Life Expectancy, Plan Cache Hit Ratio

Reading from memory much faster than reading form disk.

Not rebooting unless absolutely needed, Install updates less often (every 4 to 6 weeks common)

TempDB Performance

• What is TempDB, What uses TempDB

• Performance Recommendations

What is in TempDB?

• Temporary user objects that are explicitly created, such as: • Global or local temporary tables

• Temporary stored procedures

• Table Variables

• Sorts (including DISTINCT)

• Hash (aggregates and joins)

• Spools (table and index)

• Row versions that are generated by data modification transactions in a database that uses read-committed using row versioning isolation or snapshot isolation transactions.

• Centricity Upgrades

TempDB is Special

• No backups / No restores

• Simple recovery only

• Cannot change owner

• Cannot drop database

• Cannot Enable CDC / Mirroring / Snapshots / Availability Groups

• If the database files are missing when SQL Server starts, the files are recreated

TempDB Performance

•Performance Recommendations• Multiple Files

• Drive of its own

• Fast Storage

• 64K block size

• VLF Count (Virtual Log Files) - Also applies to other database log files.

Backup and Restore

• Compressed Backups

• Types:

• Full

• Incremental (i.e. differential)

• Logs

• GE Job and log backups

Backup and Restore

• VLF Impact on restore

• Less VLF’s = Faster Restore

• VM Snapshots• Timing with other backups

• Veeam Backups

• Watch out for truncated log files.

More info on VLFs at http://SteveStedman.com/vlf

Backup and Restore

• Automatic Restore Script• Save time and mistakes when you need to do a database

restore

• Your backups are only as good as what you have tested.

http://databasehealth.com/about/shared-scripts/

Maintenance Plan Problems

• Default maintenance plan order

• Sp_maint what it does

• Detecting Database Corruption

Best Practices

• IFI – Instant File Initialization http://stevestedman.com/IFI

• Virtualization – Don’t Overcommit

• Multiple drives (data, log, tempdb)

• Storage Medium

• Document management – Export images

• Sp_spaceused ‘patientprofileattachment’

More Best Practices

• Autogrowth Percentage and size

• Avoid autoshrink on database.

• Email configuration• Email on jobs failing

• Email on alerts.

• Changing database owners – don’t do this for Centricity.

Daily Checklist

• Check Backups

• Check the SQL Error Logs

• Quick Scan Report –Database Health Monitor

• Check scheduled jobs

Next Steps

• Take a look at the white paper. http://centricitydatabasehelp.com/

• Download the slides from this session. http://SteveStedman.com/Chug2016

• Download Database Health Monitor. http://DatabaseHealth.com

• Evaluate your system.

• Take base line vitals on your system.

• Follow up if you have any questions.

• SStedman@sqldatapartners.com

Any Questions

Contact Us