Once upon a time, somewhere in a dark, lonely data center room an SQL Server stopped talking. Nothing about it in monitoring system, except - connection lost. Ping, Telnet, Remote Login doesn't work. In the same time somewhere in the City one of the millions mobiles can't stop ringing and ringing. It's Your phone and You are DBA. It's 2AM and Disaster just began... In this session I want to discuss about best practices for surviving disaster, based on my experience in few years in data center. We will talk about more aspects which You probably know, and some about You have never think before THIS day - The Disaster Day.
5. 2 AM
A DISASTER JUST BEGAN…
Tobiasz Janusz Koprowski
Communnity Leader , SQL Server MVP
@KoprowskiT
6. • Polish SQL Server User Group Leader
• Microsoft Certified Trainer
• PASS SQL Azure Virtual Chapter Co-Founder
• Blogger, Influencer, Technical Writer
• Last 8 years living in Data Center in Wrocław
• Generally about 14 years in IT/banking area
• Trainer, Auditor, Consultant
• Speaker at SQL Server Community Launch, Time for SharePoint,
CodeCamps, SharePoint Community Launch, CISSP Day, SQL in
the City, InfoTRAMS, SQL Bits, SQL Saturday, CareerCon,
Sharepoint & SQL Connection, IT Camp,
• Deep Dives Co-Author:
High availability of SQL Server in the context
of Service Level Agreements (Chapter 18th)
ABOUT ME
7. • Your best time for dreaming … is the best
time for Disaster
• Your mobile phone ringing and ringing…
• And Your husband / wife say…….
2:00 AM … In a dreams…
8. • What’s happen with my server?
• When I made last backup?
• Where is my backup?
• Have I ever tried to restore?
– If yes – I hope that all in a team (about team
soon) remember about (me?)
– If not – who can help me NOW?
2:15 AM … in a car
9. • $#$$@$^^#^&^@!#
• Is Windows Server alive?
– YES (thanks all saints)
– or NOT (damn)
– who is responsible for it?
• Is my SQL Server alive?
– YES (why phones ringing)
– or NOT (…)
– Why I’m responsible for it?
2:40 AM … in a SERVER ROOM
13. ITS ONLY ONE: BE PREPARED
Backups
• about type of backup ( simple rm / full rm)
• about place for stored backup data
• about backup window
• about procedure used for backup
• about backup tools
• about backup of „backup logs”
• about estimated time for executing backup
• about REAL TIME of executing backup
BEST PRACTICE FOR SURVIVING DISASTER
14. In the request, backup, should include the following information:
• Information about the operating system and application version for online
backup and installed updates for these components
• a file backup policy, in particular:
a number of versions of a file stored
the storage time of the next version of the file
the frequency of execution of such incremental backups with the
proposal of their implementation
• Online Backup Policy
the storage time of a full backup with storage time such an
incremental backup
the time of transaction log files
the frequency of execution of a full backup with the proposal deadline for
its implementation
the frequency of execution of the transaction log backup
• Information about trees directories / files that should be omitted
or included during backup tasks (include / exclude list)
• Number and type / model of physical processors,
• Does the node will use the connection to the SAN to implement backup
BACKUP > extract from SOP*
15. This register contains information about a backup plan implemented.
• Backup file space:
– number of versions of a file stored in a backup
– number of days that are kept more versions of a file
– number of versions of a file stored in the backup system after its removal from client device
– number of days that will store the latest version of the deleted file from the client device
– number of days that will be stored in the archive
• Online Backup:
– number of backups stored in full backup
– number of incremental backups / diff / full stored in the backup
– frequency of transaction log backups stored in the backup (for databases)
– number of days the backups to be stored on-line backup system
• The list of nodes defined in the system backup:
– Domain | Node name | IP address of the node
– The list of defined backup tasks (called schedule)
– name of the task (schedule) | execution time
– a period of at which the task is repeated
BACKUP (registry) > extract from SOP*
16. BEST PRACTICES BY BRENT OZAR
SQL Server Backup Best Practices |
Written on October 17, 2007 by Brent Ozar in SQL Server
>> http://bit.ly/12oXm4h
I’ve been backing up SQL Servers for almost a decade now, and it’s time to share the lessons
I’ve learned. All of this is my own opinion – your mileage may vary – but I’ll try to explain the
reasoning behind the choices I make. I won’t address log shipping or snapshots this time
around.
• Never back up databases to local disk.
• Back up databases to a fileshare, then back the share up to tape.
• Cost justify the network share with lower licensing costs & simpler backups.
• Back up to a different SAN if possible.
• My sweet spot for the backup array is raid 10 SATA.
• Backup agents like NetBackup and Backup Exec mean giving up scheduling control.
• Do regular fire drill rebuilds and restores.
• Build a standalone restore testbed.
• Keep management informed on restore time estimates.
• Trust no one.
17. ITS ONLY ONE: BE PREPARED
Restore
• about type of backup ( simple rm / full rm)
• about place for stored backup data
• about the procedures of recovery
• about estimated time for recovery
• about REAL TIME for recovery
• about tools for recovery
• about Corporate Backup Manager
• about password for access to library
BEST PRACTICE FOR SURVIVING DISASTER
18. Register for Recovery/Restore/Replacement Tests
This register contains information about the tests and replacement of part or
all of the environment. It consists of the following fields:
• the date of commencement and completion of the recovery test
• client for which the test was performed recovery test
• servers involved in testing and replacement
• extent of testing and replacement
• person / persons performing the recovery test
• person on the client side accepts the correctness of the recovery test
• subsequent to the recovery test
RESTORE > extract from SOP*
19. ITS ONLY ONE: BE PREPARED
Procedures
• It is not about stored procedures!!!
• It’s about storing procedures with answers for the following:
– One piece of paper
– How to start restore
– Who can help
– How to processing a restore
– When we can finish
• It MUST be simple
BEST PRACTICE FOR SURVIVING DISASTER
20. ITS ONLY ONE: BE PREPARED
Roles
• Database Administrator
• Windows Administrator
• Backup Administrator
• Network Administrator
• Customer Key Account
• Manager of division
• Data Center Manager
• Nightshift Operator - BOFH
• Customer Administrator!!
BEST PRACTICE FOR SURVIVING DISASTER
21. ITS ONLY ONE: BE PREPARED
PSO > USO > SLA
• PSO Planned System Outages – Planned System Unavailability
– Minimum planned unavailability, due to the need to carry out
modernization work, installing patches, replacement / extension
of hardware,
– Agreed/accepted by/with the client and not affecting the
provisions of the HA, and SLA, until
• ...USO Unplaned System Outages – Unplanned System
Unavailability
– an error that prevents a partial or total work environment in a
tangible, measurable customer
– resulting in high costs if you need repairs, as well as penalty
payments for non-SLA
BEST PRACTICE FOR SURVIVING DISASTER
22. The Magic nines…
Availability %
Downtime per
year
Downtime per
month*
Downtime per
week
90% 36.5 days 72 hours 16.8 hours
95% 18.25 days 36 hours 8.4 hours
98% 7.30 days 14.4 hours 3.36 hours
99% 3.65 days 7.20 hours 1.68 hours
99.5% 1.83 days 3.60 hours 50.4 min
99.8% 17.52 hours 86.23 min 20.16 min
99.9% ("three nines") 8.76 hours 43.2 min 10.1 min
99.95% 4.38 hours 21.56 min 5.04 min
99.99% ("four nines") 52.6 min 4.32 min 1.01 min
99.999% ("five nines") 5.26 min 25.9 s 6.05 s
99.9999% ("six nines") 31.5 s 2.59 s 0.605 s
27. ITS ONLY ONE: BE PREPARED
Envelope
With ACTUAL!!! User names and passwords for:
• Windows Server Administrator
• SQL Server Administrator
• SQL Server Agent
• SQL Server Services (if You didn’t use default)
• SQL Server Applications Services
• Backup accounts
BEST PRACTICE FOR SURVIVING DISASTER
28. ITS ONLY ONE: BE PREPARED
Hardware
Some of the hard stuff for replacement:
• Server
• Motherboard
• Memory (RAM)
• Processor (CPU)
• Network Adapter (LAN/NIC)
• Fibre Channel Adapter
• Hard Disk (IDE/SATA/SAS/SSD…)
• RAID Controller
BEST PRACTICE FOR SURVIVING DISASTER
29. ITS ONLY ONE: BE PREPARED
Software
• Windows
– 2000/2003/2003R2/2008/2008R2/2012
– SP 1,2,3,4 +CU 1, 2,3, …
– Standard, Enterprise, Data Center
– x32, x64, ia64
• SQL Server
– 6.5, 7.0, 2000, 2005, 2008, 2008R2, 2012
– SP 1,2,3,4 +CU, 1,2,3,4,5,6,7,8,9,10,11,12,13….
– Drivers (servers, lan card, video card)
• AGENT ORANGE
BEST PRACTICE FOR SURVIVING DISASTER
30. ITS ONLY ONE: BE PREPARED
Keys
Some keys which You need…
• Serial keys
• Rack keys
• Server keys
• Storage keys
• Knife
• Torchlight; torch; flashlight (lighter)
• Phone
BEST PRACTICE FOR SURVIVING DISASTER
31. ITS ONLY ONE: BE PREPARED
ENCRYPTION
If You use encryption (such a TDE)
• TDE
– Create encryption key
– Export encryption key
– Backup encryption key
• CA
– Remember about expiration date
• BitLocker
BEST PRACTICE FOR SURVIVING DISASTER
32. ITS ONLY ONE: BE PREPARED
TEAM
You can work with disaster as:
Team Member
Team Leader
Last Samurai
BEST PRACTICE FOR SURVIVING DISASTER
33. ITS ONLY ONE: BE PREPARED
MANAGERS
hmm
BEST PRACTICE FOR SURVIVING DISASTER
51. ITS ONLY ONE: BE PREPARED
• Backups (and know-how about stored place, and restoring way)
• Procedures (the short is better | one page is the best)
• Roles (who can help, who is necessary for access)
• SLA (90? 95? 99,99? in minutes, hours or days you have to recover)
• Envelope (with user names and passwords for all important
accounts)
• Hardware (server, motherboard, CPU, RAM, LAN, HDD, SDD, USB)
• Software (Windows+SP+CU, SQL+SP+CU, drivers, AGENT ORANGE)
• Keys (serial numbers, physical keys, knife)
• ENCRYPTION (arrghhhhh!!! Certificates, keys, internal/external)
• TEAM (Team, leader, separate…)
• MANAGERS (hmmm)
BEST PRACTICE FOR SURVIVING DISASTER